Hitachi Command Suite Dynamic Link Manager (for Solaris) User Guide Document Organization Product Version Getting Help Contents MK-92DLM114-28
© 2014 Hitachi, Ltd. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying and recording, or stored in a database or retrieval system for any purpose without the express written permission of Hitachi, Ltd. Hitachi, Ltd., reserves the right to make changes to this document at any time without notice and assumes no responsibility for its use.
Contents Preface.................................................................................................xiii Intended audience...................................................................................................xiv Product version....................................................................................................... xiv Release notes..........................................................................................................xiv Document revision level....
Path Status Transition..................................................................................... 2-19 The Online Path Status.......................................................................... 2-19 The Offline Path Status.......................................................................... 2-20 Status Transitions of a Path....................................................................2-20 Intermittent Error Monitoring (Functionality When Automatic Failback Is Used).........
Notes on Hardware Settings............................................................................ 3-39 Notes on Installation....................................................................................... 3-40 Notes on Related Software.............................................................................. 3-44 Notes on Command Execution......................................................................... 3-45 Notes on the Disk Label.....................................................
Creating a New Boot Environment........................................................ 3-112 Configuring the Post-Migration ZFS Boot Disk Environment..................... 3-113 Migrating to the ZFS Boot Disk Environment.......................................... 3-114 Performing a Check after Restart.......................................................... 3-115 Migrating from a Boot Disk Environment to the Local Boot Disk Environment...........
Devices to Be Removed from VxVM.......................................................3-161 Removing Devices from VxVM on a Controller Basis............................... 3-162 Removing Devices From VxVM on a Path Basis...................................... 3-166 Actions To Be Taken if an sd or ssd Device Has Not Been Suppressed from VxVM..................................................................................................3-170 Introducing VxVM while Using HDLM..................................
Initializing HDLM When the Host Is Started in Single-User Mode.................4-4 Tasks that Can Be Performed in Single-User Mode..................................... 4-5 Maintenance Tasks on Devices Connected by Paths in the Boot Disk Environment .......................................................................................................................4-6 HDLM Operations Using Commands......................................................................... 4-6 Notes on Using Commands.........
Specifying Whether a Logical Unit Is To Be Managed by HDLM (When Using Solaris Cluster)..........................................................................................................4-52 Changing an HDLM-Managed Device to a Non-HDLM-Managed Device...... 4-52 Changing a Non-HDLM-Managed Device to an HDLM-Managed Device (When the Node Must Be Restarted).................................................................
online (Places Paths Online)...................................................................................6-12 Format...........................................................................................................6-12 To place paths online.............................................................................6-12 To display the format of the online operation.......................................... 6-12 Parameters........................................................................
To Display the Format of the delete Operation.........................................6-81 Parameters.....................................................................................................6-81 To Delete a Path Dynamically................................................................. 6-81 To Display the Format of the delete Operation.........................................6-82 7 Utility Reference...................................................................................
Format and Meaning of Message IDs................................................................. 8-3 Terms Used in Messages and Message Explanations............................................8-3 Components That Output Messages to Syslog.....................................................8-3 KAPL01001 to KAPL02000....................................................................................... 8-4 KAPL03001 to KAPL04000................................................................................
Preface This document describes how to use the Hitachi Dynamic Link Manager.
Intended audience This document is intended for storage administrators who use Hitachi Dynamic Link Manager (HDLM) to operate and manage storage systems, and assumes that readers have: • Knowledge of Solaris and its management functionality • Knowledge of Storage system management functionality • Knowledge of Cluster software functionality • Knowledge of Volume management software functionality Product version This document revision applies to HDLM for Solaris version 8.0.0 or later.
Document organization The following table provides an overview of the contents and organization of this document. Click the chapter title in the left column to go to that chapter. The first page of each chapter provides links to the sections in that chapter. Chapter/Appendix Description Chapter 1, Overview of HDLM on page 1-1 Gives an overview of HDLM, and describes its features.
• Hitachi Adaptable Modular Storage Series User's Guide • Hitachi Simple Modular Storage Series User's Guide • Hitachi Unified Storage Series User's Guide • Hitachi USP Series User's Guide • Hitachi Workgroup Modular Storage Series User's Guide • Thunder9580V Series Disk Array Subsystem User's Guide • Hitachi Universal Storage Platform V Hitachi Universal Storage Platform VM User and Reference Guide • Hitachi Virtual Storage Platform Series User's Guide • Reference Manual / File Conversion
Physical capacity unit Value 1 kilobyte (KB) 1,000 (103) bytes 1 megabyte (MB) 1,000 KB or 1,0002 bytes 1 gigabyte (GB) 1,000 MB or 1,0003 bytes 1 terabyte (TB) 1,000 GB or 1,0004 bytes 1 petabyte (PB) 1,000 TB or 1,0005 bytes 1 exabyte (EB) 1,000 PB or 1,0006 bytes Logical storage capacity values (for example, logical device capacity) are calculated based on the following values: Logical capacity unit Value 1 block 512 bytes 1 KB 1,024 (210) bytes 1 MB 1,024 KB or 1,0242 bytes 1 GB 1
our existing Support Portal and support services by providing an area where you can get answers to non-critical issues and questions. Join the conversation today! Go to community.hds.com, register, and complete your profile. Comments Please send us your comments on this document: doc.comments@hds.com. Include the document title and number, including the revision level (for example, -07), and refer to specific sections and paragraphs whenever possible.
1 Overview of HDLM HDLM is a software package that manages paths between a host and a storage system. HDLM is designed to distribute loads across multiple paths and will switch a given load to another path if there is a failure in the path that is currently being used, thus improving system reliability. This chapter gives an overview of HDLM and describes its features.
What is HDLM? With the widespread use of data warehousing and increasing use of multimedia data, the need for high-speed processing of large volumes of data on networks has rapidly grown. To satisfy this need, networks dedicated to the transfer of data, such as SANs, are now being used to provide access to storage systems. HDLM manages the access paths to these storage systems.
The ability to continue running operations between a host and storage system, even if there is a failure. This is also known as performing a failover. When a host is connected to a storage system via multiple paths, HDLM can automatically switch to another path if there is some sort of failure in the path that is currently being used. This allows operations to continue between a host and a storage system.
1-4 Overview of HDLM Hitachi Dynamic Link Manager (for Solaris) User Guide
2 HDLM Functions This chapter describes the various functions that are built into HDLM. Before the function specifications are explained though, this chapter will go into detail about the HDLM management targets, system configuration, and basic terms that are necessary to know to effectively operate HDLM. After that, the rest of the chapter focus on describing all the HDLM functions, including the main ones: load distribution across paths and path switching.
□ Collecting Audit Log Data □ Integrated HDLM management using Global Link Manager □ Cluster Support 2-2 HDLM Functions Hitachi Dynamic Link Manager (for Solaris) User Guide
Devices Managed by HDLM Below is a list of devices that can or cannot be managed by HDLM. The devices that can be managed by HDLM are called HDLM management-target devices. HDLM management-target devices: The following devices are from the storage systems listed in Section What is HDLM? on page 1-2: ¢ SCSI devices (sd or ssd devices) ¢ Boot disks# ¢ Swap devices# ¢ Dump devices# #: If you want to use these disks as HDLM management-target devices, assign VTOC labels to them.
paths. The ID that HDLM assigns to each path (physical path) is called a AutoPATH_ID. A path is also sometimes called a managed object. The following figure shows the HDLM system configuration. Figure 2-1 HDLM System Configuration The following table lists and describes the HDLM system components. Table 2-1 HDLM System Components Components 2-4 Description HBA A host bus adapter. This serves as a cable port on the host.
LU Configuration After you have properly installed HDLM, the LU configuration will change as follows: Before the installation of HDLM: The host recognizes that an sd or ssd device is connected to each physical path. Thus, a single LU in the storage system is recognized as the same number of LUs as that of physical paths. After the installation of HDLM: An HDLM device that corresponds one-to-one with the Dev in an LU in the storage system is created above an sd or ssd device.
Figure 2-2 LU Configuration Recognized by the Host After the Installation of HDLM The following table lists and describes the components recognized by the host. Table 2-2 Components Recognized by the Host Components Description HLU An LU that the host recognizes via the HDLM driver. It is called a host LU. No matter how many physical paths exist, one host LU is recognized for one LU in the storage system. HDev A Dev (a slice or partition) in an LU that the host recognizes via the HDLM driver.
Figure 2-3 Configuration of the HDLM Programs The following table lists and describes the functions of these programs.
Program name Functions • Request path health checks and automatic failbacks to be performed • Collect error log data HDLM alert driver Reports the log information collected by the HDLM driver to the HDLM manager. The driver name is dlmadrv. HDLM driver Controls all the HDLM functions, manages paths, and detects errors. The HDLM driver consists of the following: • Core logic component • Filter component Controls the basic functionality of HDLM. Sends and receives I/O data.
Figure 2-4 Position of the HDLM Driver and HDLM Devices Logical Device Files for HDLM Devices When you install HDLM, a logical device file to be used by HDLM will be created for each LU on a per-Dev (slice) basis. Setting this logical device file name in an application, such as volume management software, enables the application to access an LU by using the HDLM function. The logical device files existing before HDLM installation (the logical device files of an sd or ssd) will be deleted.
and for one of the Dev (slices) in that LU, the corresponding logical device file names of the sd or ssd devices are c2t1d1s0 and c3t2d1s0. In this case, when you install HDLM, these logical device files will be deleted. Then, a logical device file that has a different controller number, such as c4t1d1s0, is created for the HDLM device.
Figure 2-5 Flow of I/O Data When the Load Balancing Function Is Not Used When the load balancing function is not used, I/O operations converge onto a single path (A). The load on that one physical path (A) will cause a bottleneck, which might cause problems with system performance.
Figure 2-6 Flow of I/O Data When the Load Balancing Function Is Used When the load balancing function is used, I/O operations are distributed via multiple physical paths (A, B, C, and D). This helps to prevent problems with system performance and helps prevent bottlenecks from occurring. Paths to Which Load Balancing Is Applied This section describes, for each type of storage system, the paths to which the load balancing function is applied.
remaining usable owner paths. When all owner paths cannot be used, load balancing is performed among the non-owner paths. For the example in Figure 2-7 Overview of Load Balancing on page 2-13, suppose that in the owner controller of LU0 is CHA0. When the LU is accessed, the load is balanced between the two physical paths A and B, which are both owner paths. When one of the paths (A) cannot be used, then the LU is accessed from the only other owner physical path (B).
¢ Hitachi AMS2000 series# ¢ Hitachi SMS series# ¢ HUS100 series# HUS VM #: This storage system applies when the dynamic I/O path control function is disabled.
The paths are simply selected in order from among all the connected paths. ¢ ¢ Extended Least I/Os The path that has the least number of I/Os being processed is selected from among all the connected paths. Extended Least Blocks The path that has the least number of I/O blocks being processed is selected from among all the connected paths. The following table lists and describes the features of the load balancing algorithms.
If a path in which an error has occurred recovers from the error, HDLM can then switch back to that path. This is called a failback. Two types of failovers and failbacks are available: • Automatic failovers and failbacks • Manual failovers and failbacks Failovers and failbacks switch which path is being used and also change the statuses of the paths. A path status is either online or offline. An online status means that the path can receive I/Os.
For example, in Figure 2-8 Path Switching on page 2-17, the owner controller of LU is CHA0, and access to the LU is made only via the physical path (A). After the access path is placed offline, the first candidate for the switching destination is the physical path (B) and the second candidate is the physical path (C or D).
Automatic Failbacks When a path recovers from an error, HDLM can automatically place the recovered path back online. This function is called the automatic failback function. In order to use the automatic failback function, HDLM must already be monitoring error recovery on a regular basis. When the Thunder 9500V series, or Hitachi AMS/WMS series is being used, HDLM selects the path to use from online owner paths, and then from online non-owner paths.
However, if there is only one online path for a particular LU, that path cannot be manually switched offline. Also, a path with an error that has not been recovered from yet cannot be switched online. HDLM uses the same algorithms to select the path that will be used next, regardless of whether automatic or manual path switching is used. When the Thunder 9500V series, or Hitachi AMS/WMS series is being used, HDLM selects the switching destination path from owner paths and then from non-owner paths.
If none of the paths accessing a single LU are in the Online status, one of the paths will be changed to the Online(E) status. All the paths that access the same LU will not be in the offline statuses. This ensures access to the LU. The text (E) of Online(E) indicates the error attribute, which indicates that an error occurred in the path. • Online(S)# The paths to the primary volume (P-VOL) in the HAM environment have recovered from an error, but I/O to the P-VOL is suppressed.
Figure 2-9 Path Status Transitions Legend: Online operation: Online operation performed by executing the dlnkmgr command's online operation. Offline operation: Offline operation performed by executing the dlnkmgr command's offline operation. #1 When no Online or Offline(E) paths exist among the paths that access the same LU.
#4 One of the Offline(E) paths is changed to the Online(E) path. #5 When an Offline(E) path exists among the paths that access the same LU. Figure 2-10 Path Status Transitions (P-VOL in HAM environment) Legend: Online operation: Online operation performed by executing the dlnkmgr command's online operation. Offline operation: Offline operation performed by executing the dlnkmgr command's offline operation. #1 Also when an error occurs in all the paths to an S-VOL in the Online(D) status.
When you are using intermittent error monitoring, the path in which the intermittent error occurred is not automatically placed online even when the path recovers from the error. In such a case, place the path online manually. Note If there is a path failure immediately after a path is placed offline by using the dlnkmgr command, Offline(C) might change to Offline(E).
determined to have an intermittent error if an error occurs 3 or more times in 30 minutes. You can set up intermittent error monitoring by executing the dlnkmgr command's set operation. Intermittent error monitoring can be used only when automatic failback has already been enabled. The values that can be specified for intermittent error monitoring depend on the values specified for automatic failbacks. For details on how to specify the settings, see set (Sets Up the Operating Environment) on page 6-17.
Figure 2-11 Action What Will Happen When an Intermittent Error Occurs on a Path When an Intermittent Error Does Not Occur If an error does not occur in the path the specified number of times within the specified interval, the system determines that the path does not have an intermittent error. In such a case, the error monitoring will finish when the specified error-monitoring interval finishes, upon which the number of errors is reset to 0.
As shown in Figure 2-12 What Will Happen When an Intermittent Error Does Not Occur on a Path on page 2-25, normally, the count for the number of times that an error occurs is started after the path is first recovered from an error by using the automatic failback function.
When a User Changes the Intermittent Error Information The following might be reset when a user changes any of the values set for the intermittent error or the path status: the number of errors that have already been counted during error monitoring, the amount of time that has passed since error monitoring has started, and the information about whether an intermittent error has occurred.
User operation Number of errors and time passed since error monitoring started monitoring is being performed Information about paths not subject to automatic failback If a path has been removed from the paths subject to automatic monitoring, that path is no longer monitored. Restarting the HDLM manager Reset#3 Inherited Restarting the host Reset Reset #1 When you disable the intermittent error monitoring function, information about paths not subject to automatic failback will be reset.
connected to non-owner paths, we recommend that you use path health checking to detect errors. This enables the system to use the most up-to-date path-status information when selecting the next path to use. You can configure path health checking by executing the dlnkmgr command's set operation. For details on the set operation, see set (Sets Up the Operating Environment) on page 6-17.
Error Management For troubleshooting purposes, HDLM collects information and stores it into log files. The error information to be collected can be filtered out by error level, and then stored into the log files. The following figure shows the flow of data when error information is collected on a host which is running HDLM. Figure 2-14 Flow of Data When Collecting Error Information Logs might be collected in layers below HDLM, such as for the SCSI driver. For more details, see the Solaris documentation.
Types of Collected Logs HDLM collects information on detected errors and trace information in the integrated trace file, trace file, error logs, log for the dlmcfgmgr utility for managing the HDLM configuration, and syslog. You can use the error information to examine the status of an error and analyze the cause of the error. The following table lists and describes the error information that can be collected in logs.
Log name Description file /etc/syslog.conf or /etc/ rsyslog.conf are collected. # We recommend that you configure the system so that information at the Information level and higher is output. Output destination The syslog file path is specified in the file /etc/ syslog.conf or /etc/ rsyslog.conf. For details, refer to the Solaris documentation. Syslogs can be checked using a text editor.
In syslog, the HDLM messages on and above the level set by the user configured in /etc/syslog.conf or /etc/rsyslog.conf are collected. It is recommended that you set the information to be output at the info level and higher. Note that the facility and level (in facility format) when HDLM outputs messages to syslog are all user. The error information in error logs and trace files are collected based on a user-defined collection level.
following table describes the categories of audit log data that Hitachi storagerelated products can collect.
Category Maintenance AnomalyEvent Explanation • Updating account settings, such as adding and deleting accounts • Setting up security • Viewing or updating audit log settings An event indicating that a maintenance operation has terminated normally or failed, including: • Adding or removing hardware components • Adding or removing software components An event indicating an abnormal state such as exceeding a threshold, including: • Exceeding a network traffic threshold • Exceeding a CPU load t
Category Authentication ConfigurationAccess 2-36 Explanation Administrator or end-user authentication Viewing or updating configuration information Audit event Severity# 1 Message ID Startup of the I/O 6 information monitoring function was successful. KAPL15112-I Startup of the I/O 3 information monitoring function failed. KAPL15113-E The I/O 6 information monitoring function stopped. KAPL15114-I The I/O 4 information monitoring function terminated.
Category Explanation Audit event Severity# 1 Message ID environment was successful. Setup of the 3 operating environment failed. KAPL15106-E An attempt to display program information was successful. 6 KAPL15107-I An attempt to display program information failed. 3 KAPL15108-E An attempt to display HDLM managementtarget information was successful. 6 KAPL15109-I An attempt to display HDLM managementtarget information failed.
Category Explanation Audit event Severity# 1 Message ID d] command was successful. Processing of the dlmvxexclude [d] command failed. 3 KAPL15027-E The status of a path was successfully changed to Online. 6 KAPL15116-I A path was 6 successfully added. KAPL15117-I Path addition failed.
¢ If the severity specified by the HDLM command's set operation differs from the severity specified by the configuration file /etc/syslog.conf or /etc/rsyslog.conf, the higher severity level is used for outputting audit log data. Destination and Filtering of Audit Log Data Audit log data is output to syslog. Because HDLM messages other than audit log data are also output to syslog, we recommend that you specify the output destination that is used exclusively for audit log data.
¢ All of the above For details on how to specify audit log settings, see Setting Up the HDLM Functions on page 3-140. Audit Log Data Formats The following describes the format of audit log data: Format of audit log data output to syslog: ¢ priority ¢ date-and-time ¢ host-name ¢ program-name ¢ [process-ID] ¢ message-section The following shows the format of message-section and explains its contents.
Item# Explanation Audit event type Event type Audit event result Event result Subject ID for audit event result Depending on the event, an account ID, process ID, or IP address is output.
For large-scale system configurations using many hosts running HDLM, the operational load for managing paths on individual hosts increases with the size of the configuration. By linking HDLM and Global Link Manager, you can centrally manage path information for multiple instances of HDLM and reduce operational load. In addition, you can switch the operational status of paths to perform system-wide load balancing, and centrally manage the system by collecting HDLM failure information in Global Link Manager.
For details on cluster software supported by HDLM, the supported Solaris version, and usable volume management software, see Combinations of Cluster Software and Volume Managers Supported by HDLM on page 3-7. HDLM uses a path of the active host to access an LU. Paths are switched in units of physical paths. Therefore, if an error occurs in a path, all the other paths that run through the same physical path are switched. The switching destination is a physical path of the active host.
2-44 HDLM Functions Hitachi Dynamic Link Manager (for Solaris) User Guide
3 Creating an HDLM Environment This chapter describes the procedures for setting up an HDLM environment and the procedure for canceling those settings. Make sure that HDLM is installed and the functions have been set up. Set up volume groups and cluster software to suit your operating environment.
□ Creating File Systems for HDLM (When Volume Management Software Is Not Used) □ Setting Up VxVM □ Setting Up SDS □ Setting Up SVM □ Setting Up VCS □ Removing HDLM 3-2 Creating an HDLM Environment Hitachi Dynamic Link Manager (for Solaris) User Guide
HDLM System Requirements Check the following before installing HDLM: For the requirements for using HDLM in an HAM environment, see the release notes of HDLM. Hosts and OSs Supported by HDLM HDLM can be installed on a SPARC series computer which is running an OS listed in the following table.
Table 3-2 JDK required for linking with Global Link Manager OS Solaris 8 or Solaris 9 Solaris 10 Solaris 11 JDK package • JDK 1.4.2_15 or later (32 bit version) • JDK 5.0_11 or later (32 bit version) • JDK 6.0_17 or later (32 bit version) • JDK 1.4.2_15 or later (32 bit version) • JDK 5.0_11 or later (32 bit version) • JDK 6.0_17 or later (32 bit version) • JDK 7.0 (32 bit version) • JDK 6.0_17 or later (32 bit version) • JDK 7.
The applicable storage systems require a dual controller configuration. If you use the system in a hub-connected environment, you must set unique loop IDs for all connected hosts and storage systems. For details on the micro-programs and settings information for storage systems, which are required to use HDLM, see the HDLM Release Notes and maintenance documentation for the storage system.
Cluster Software Supported by HDLM The following table lists the cluster software versions supported by HDLM when building a cluster configuration. Table 3-3 Supported cluster software versions Cluster software Supported versions Local boot disk environment#1 Boot disk environment#2 Solaris Cluster#5 Sun Cluster 3.1, Sun Cluster 3.2, Oracle Solaris Cluster 3.3, or Oracle Solaris Cluster 4.0 Sun Cluster 3.1 8/05 (Update 4) Oracle RAC#6 Oracle9i RAC, Oracle RAC 10g, or Oracle RAC 11g -- VCS#3 VCS5.
• When combining configurations by using VxVM VxVM 4.1#1 or VxVM 5.0#1 HDLM-managed boot disks do not support a mirrored boot disk configuration incorporating a volume manager such as SDS or VxVM. For this reason, you cannot register an HDLM-managed boot disk in bootdg when using VxVM.
OS Cluster Sun Cluster 3.1 (9/04)#1 Volume manager None SDS 4.2.1 Sun Cluster 3.1 (8/05)#1 None SDS 4.2.1 VCS 5.0#2 #3 None VxVM 5.0 Solaris 9 None None SVM#4 VxVM 5.0#5 Sun Cluster 3.1#1 #5 None SVM#6 Sun Cluster 3.1 (9/04)#1 #5 None SVM#6 Sun Cluster 3.1 (8/05)#1 #5 None SVM#6 VCS 5.0#2 #3 #5 None VxVM 5.0 Solaris 10 None None SVM 1.0#4 #7 #8 VxVM 5.0#5 #8 Sun Cluster 3.1 (8/05) #1 #5 #8 None SVM#6 #9 Sun Cluster 3.2 #1 #10 None#11 SVM#5 #6 #8 #9 VxVM 5.0#5 #8 #12 Sun Cluster 3.
OS Cluster Volume manager SVM #4 #9 VxVM 5.0 Oracle Solaris Cluster 3.3#10 #13 None SVM#5 #6 #8 #9 VxVM 5.1#5 #8 VCS 5.0#3 None#14 VxVM 5.0#5 #8 #15 VCS 5.1#14 None VxVM 5.1#5 #8 Solaris 11 VCS 6.0#16 VxVM 6.0#5 #8 Oracle Solaris Cluster 4.0 None SVM VCS 6.0 VxVM 6.
¢ Multi-owner disksets ¢ Diskset import ¢ Automatic (top down) volume creation #7 In a configuration that uses a driver other than the Oracle HBA driver (other than the qlc or emlxs driver), the SVM shared diskset cannot use disks managed by HDLM. #8 Does not support ZFS.
The I/O fencing function can be used only when Hitachi USP series, Universal Storage Platform V/VM series, or Virtual Storage Platform series storage systems are connected in a Solaris 10 environment. Note that the only supported service group type is the failover service group. The parallel service group and hybrid service group are not supported. When Creating an Oracle9i RAC Environment Required programs The following table lists the programs required to create an Oracle9i RAC environment.
Table 3-6 Programs required to create an Oracle RAC 10g environment (For Solaris 8 or Solaris 9) Configurati on Program Configuration OS 1 Solaris 8 or Solaris 9 Remarks • In Solaris 8, use Update 7 or later. • In Solaris 9, use Update 6 or later. Oracle RAC 10g Oracle 10g Database 10.1.0.2.0 -- Cluster Oracle Cluster Ready Services (CRS) 10.1.0.2.0 -- Volume Manager ASM • ASM is bundled with Oracle RAC 10g.
Configurati on Program Remarks the documentation for Oracle RAC 10g. Configuration OS Solaris 9 3 Oracle RAC 10g Oracle 10g Database 10.1.0.4.0 --- Cluster Oracle Clusterware 10.1.0.4.0 -- Volume Manager None (Specify an HDLM raw device) -- Configuration OS Solaris 9 4 Oracle RAC 10g Oracle 10g Database 10.1.0.5.0 --- Cluster Oracle Clusterware 10.1.0.5.0 -- Volume Manager None (Specify an HDLM raw device) -- Configuration OS Solaris 9 5 Oracle RAC 10g Oracle 10g Database 10.2.0.2.
Configurati on Program Remarks of the Oracle database. In Oracle RAC 10g, HDLM devices can be used following the same procedures as for disk devices. For details on how to install ASM, refer to the documentation for Oracle RAC 10g. Table 3-7 Programs required to create an Oracle RAC 10g environment (For Solaris 10) Configurati on Configuratio n1 Program Remarks OS Solaris 10 -- Oracle RAC 10g Oracle 10g Database 10.1.0.4.0 -- Cluster Oracle Cluster Ready Services (CRS) 10.1.0.4.
Configurati on Program Remarks Cluster Oracle Clusterware 10.2.0.1.0 -- Volume Manager ASM • ASM is bundled with Oracle RAC 10g. • ASM is used as the disk memory area for files and recovery files of the Oracle database. In Oracle RAC 10g, HDLM devices can be used following the same procedures as for disk devices. For details on how to install ASM, refer to the documentation for Oracle RAC 10g. Configuratio n4 OS Solaris 10 -- Oracle RAC 10g Oracle 10g Database 10.2.0.2.
Configurati on Configuratio n7 Configuratio n8 Configuratio n9 3-16 Program Remarks Oracle RAC 10g Oracle 10g Database 10.2.0.2.0 Only two-node configurations are supported. Cluster Sun Cluster 3.1 8/05 and Oracle Clusterware 10.2.0.2.0 Volume Manager None (Specify an HDLM raw device by the device ID of Sun Cluster) -- OS Solaris 10 -- Oracle RAC 10g Oracle 10g Database 10.2.0.2.0 Cluster Sun Cluster 3.1 8/05 and Oracle Clusterware 10.2.0.2.0 Only two-node configurations are supported.
Configurati on Program Remarks procedures as for disk devices. For details on how to install ASM, refer to the documentation for Oracle RAC 10g. Configuratio n 10 Configuratio n 11 OS Solaris 10 -- Oracle RAC 10g Oracle 10g Database 10.2.0.3.0 -- Cluster Sun Cluster 3.2 and Oracle Clusterware 10.2.0.3.0 -- Volume Manager None (Specify an HDLM raw device from the Sun Cluster device ID) -- OS Solaris 10 -- Oracle RAC 10g Oracle 10g Database 10.2.0.3.0 Cluster Sun Cluster 3.
Configurati on Program Remarks refer to the documentation for Oracle RAC 10g. Configuratio n 13 Configuratio n 14 Configuratio n 15 Configuratio n 16 3-18 OS Solaris 10 -- Oracle RAC 10g Oracle 10g Database 10.2.0.4.0 Cluster Sun Cluster 3.2 and Oracle Clusterware 10.2.0.4.0 Only two-node configurations are supported. Volume Manager VxVM 5.
Configurati on Program Remarks Oracle database files and recovery files. For the disk device used by ASM, specify the Sun Cluster device ID. For details on how to use ASM, refer to the documentation for Oracle RAC 10g. Configuratio n 17 Configuratio n 18 Configuratio n 19 OS Solaris 10 -- Oracle RAC 10g Oracle 10g Database 10.2.0.3.0 Cluster Oracle Solaris Cluster 3.3 and Oracle Clusterware 10.2.0.3.0 Only two-node configurations are supported. Volume Manager VxVM 5.
Configurati on Configuratio n 20 Program Remarks Cluster Oracle Solaris Cluster 3.3 and Oracle Clusterware 10.2.0.4.0 Volume Manager None (Specify an HDLM raw device from the Sun Cluster device ID) -- OS Solaris 10 -- Oracle RAC 10g Oracle 10g Database 10.2.0.4.0 Configurations that use Storage Foundation for Oracle RAC 5.0 (where the MP version is the same as that of VCS) are supported. Cluster VCS 5.0# The parallel service group with an I/O fencing function enabled is supported.
Target program Patch ID Timing for applying 108773-18 or later 108921-16 or later 108940-53 or later 108987-13 or later 108989-02 or later 108993-19 or later# 109147-24 or later 110386-03 or later 111023-02 or later 111111-03 or later 111308-03 or later 111310-01 or later 112396-02 or later 111721-04 or later 112003-03 or later 112138-01 or later # When using Oracle RAC 10g 10.2.0.1.0, apply 108993-45, not 108993-19.
# It is necessary only when using Oracle RAC 10g 10.1.0.4.0. Note • When a host and an Oracle RAC 10g voting disk are connected by multiple paths, HDLM performs failover processing for those paths (in the same way as for normal paths) when an I/O timeout occurs for one of the paths. Note that, depending on the settings of Oracle RAC 10g, Oracle RAC 10g might determine that a node error has occurred before the failover processing performed by HDLM is completed, and then re-configure the cluster.
• 600 seconds (default value of Oracle Clusterware) Table 3-12 Formula for Calculating MISSCOUNT Storage system type • Lightning 9900V series • Hitachi USP series • Universal Storage Platform V/VM series • Virtual Storage Platform series • VSP G1000 series • HUS VM • Hitachi AMS2000/AMS/WMS/SMS series • HUS100 series • Thunder 9500V series Formula for obtaining the value of MISSCOUNT number-of-paths-connected-to-the-voting-disk x 60 seconds number-of-paths-connected-to-the-voting-disk x
For details on how to change MISSCOUNT and DISKTIMEOUT, contact the company with which you have a contract for Oracle Support Services. Note that when you remove HDLM from the above configuration, you must reset the values of MISSCOUNT and DISKTIMEOUT to their original values. Therefore, make a note of the original values of MISSCOUNT and DISKTIMEOUT before changing them.
c. For the created alias device file, set the owner, group, and access permission mode by using the following command. The owner, group, and access permission mode to be set differs depending on the usage purpose of Oracle RAC 10g corresponding to that device. For details on this point, refer to the Oracle documentation. # chmod mode /dev/alias-device-file # chown owner:group /dev/alias-device-file Execution example: # chmod 640 /dev/crs_ocr1 # chown root:oinstall /dev/crs_ocr1 # d.
Configurati on Program Volume Manager ASM Remarks • ASM is bundled with Oracle RAC 11g. • ASM is used as the disk memory area for the Oracle database files and recovery files. In Oracle RAC 11g, HDLM devices can be used following the same procedures as for disk devices. For details on how to use ASM, refer to the documentation for Oracle RAC 11g. Configuratio OS Solaris 10 n3 Oracle RAC 11g Oracle 11g Database 11.1.0.6.0 Cluster Sun Cluster 3.1 8/05 and Oracle Clusterware 11.1.0.6.
Configurati on Program Volume Manager None (Specify an HDLM raw device) Configuratio OS Solaris 10 n6 Oracle RAC 11g Oracle 11g Database 11.1.0.6.0 Cluster Sun Cluster 3.2 and Oracle Clusterware 11.1.0.6.0 Volume Manager ASM Remarks --Only two-node configurations are supported. • ASM is bundled with Oracle RAC 11g. • ASM is used as the disk memory area for the Oracle database files and recovery files. For the disk device used by ASM, specify the Sun Cluster device ID.
Configurati on Program Remarks documentation for Oracle RAC 11g. Configuratio OS Solaris 10 n9 Oracle RAC 11g Oracle 11g Database 11.1.0.7.0 --- Cluster Oracle Clusterware 11.1.0.7.0 -- Volume Manager None (Specify an HDLM raw device) -- Configuratio OS Solaris 10 n 10 Oracle RAC 11g Oracle 11g Database 11.1.0.7.0 --- Cluster Oracle Clusterware 11.1.0.7.0 -- Volume Manager ASM • ASM is bundled with Oracle RAC 11g.
Configurati on Program Remarks Oracle RAC 11g Oracle 11g Database 11.2.0.1.0 -- Cluster Oracle Grid Infrastructure 11.2.0.1.0 -- Volume Manager ASM • ASM is bundled with Oracle Grid Infrastructure. • ASM is used as the disk memory area for the Oracle database files and recovery files. In Oracle RAC 11g, HDLM devices can be used as disk devices by following the usual procedures for HDLM devices. For details on how to use ASM, refer to the documentation for Oracle RAC 11g.
Configurati on Program Volume Manager ASM Remarks • ASM is bundled with Oracle Grid Infrastructure. • ASM is used as the disk memory area for the Oracle database files and recovery files. In Oracle RAC 11g, HDLM devices can be used as disk devices by following the usual procedures for HDLM devices. For details on how to use ASM, refer to the documentation for Oracle RAC 11g. Configuratio OS Solaris 10 n 15 Oracle RAC 11g Oracle 11g Database 11.2.0.2.0 Cluster Oracle Grid Infrastructure 11.2.0.2.
Configurati on Program Remarks recovery files. In Oracle RAC 11g, HDLM devices can be used as disk devices by following the usual procedures for HDLM devices. For details on how to use ASM, refer to the documentation for Oracle RAC 11g. Configuratio OS Solaris 11 n 17 Oracle RAC 11g Oracle 11g Database 11.2.0.3.0 Cluster Oracle Grid Infrastructure 11.2.0.3.0 Volume Manager ASM -Only two-node configurations are supported. • ASM is bundled with Oracle Grid Infrastructure.
Configurati on Program Remarks For details on how to use ASM, refer to the documentation for Oracle RAC 11g. # You must apply MP3 or later. Note • When a host and an Oracle RAC 11g voting disk are connected by multiple paths, HDLM performs failover processing for those paths (in the same way as for normal paths) when an I/O timeout occurs for one of the paths.
the change, use the following table to obtain the value to be specified, and then change the current value to a value equal to or greater than the value you have obtained.
a. Check the major number and minor number of HDLM raw devices used by Oracle RAC 11g by executing the following command in each node: # ls -lL HDLM-raw-device-file Execution example: # ls -lL /dev/rdsk/c10t50060E8005271760d5s0 crw-r----- 1 root sys 307, 1608 date/time /dev/rdsk/ c10t50060E8005271760d5s0 # In this example, the major number is 307 and the minor number is 1608. b. Create an alias device file by executing the following command in each node.
Virtualization Environments Supported by HDLM HDLM supports LDoms in a virtualization environment. LDoms consist of a control domain, and one or more service domains, I/O domains, and guest domains. The following describes domains supported by HDLM. • If the control domain and the I/O domains that manage the HBAs for HDLM-managed storage are not separated: ¢ ¢ • HDLM can be installed in the control domain.
OS LDoms Volume Manager Disk Label EFI File System None UFS VxFS ZFS ZFS Y Y N Y None Y Y N Y SVM Y Y N N ZFS Y Y N Y Legend: Y: Can be used N: Cannot be used Memory and Disk Capacity Requirements This subsection explains the memory and disk capacity requirements. Memory Requirements The host requires 11 MB of memory. Disk Capacity Requirements The following table lists the disk capacity requirements for a host.
#3 This value depends on the trace file settings. The maximum is 1000 MB. q = (t × n) ÷ 1024 (in MB), where: t is the size of the trace file (units: KB, default is 1,000), and n is the number of trace files (default is 4). The resulting fraction will be rounded up. Number of LUs and Paths Supported in HDLM The following table lists the number of LUs and paths supported in HDLM.
Figure 3-1 Flow of HDLM Environment Setup HDLM Installation Types This subsection describes the following types of HDLM installation: new installation, upgrade installation, and re-installation. New installation of HDLM: Installing HDLM in a server on which HDLM has not been installed is called a new installation of HDLM. Upgrade installation of HDLM: Installing a newer version than the existing version without removing the existing version is called an upgrade installation of HDLM.
Notes on Creating an HDLM Environment This section gives points to be noted when creating an HDLM environment. For points to be noted during operation of HDLM, see Notes on Using HDLM on page 4-2. Notes on Hardware Settings For details on the settings of the storage system, see storage system manual and maintenance documentation for the storage system. • When using HBAs that can set up a binding between target IDs and storage-system ports, you must set up the binding before installing HDLM.
Notes on Installation • You cannot edit the HDLM driver configuration definition file (/ kernel/drv/dlmfdrv.conf). To change an HDLM-managed device to a non-HDLM-managed device, use the HDLM device unconfiguration function of the HDLM device.
¢ SVS is an HDLM management-target device. You are upgrading from HDLM 5.6.1 or earlier. When both these conditions are satisfied, the previous generation of the HDLM driver configuration definition file (/kernel/drv/ dlmfdrv.conf.bak) and HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/config/dlmfdrv.unconf.bak) are also updated. For details about the dlmfdrv.conf.bak and dlmfdrv.unconf.bak files, see Note in section Parameters on page 7-25.
/var/opt/hitachi/HNTRLib2/spool/hntr2n.log (n indicates a file number) Trace files after upgrade installation or re-installation: /var/opt/DynamicLinkManager/log/hdlmtrn.log (n indicates a file number) • If you interrupted a new installation, check whether the /var/opt/ DynamicLinkManager directory exists.
For details about the commands, refer to Solaris 10 or Solaris 11 documentation. The following table shows Solaris zones supported in HDLM. Table 3-20 Support for the Solaris zones Item • Non-global zone Remarks Installing HDLM Supported Not supported -- Path operations Supported Not supported -- Displaying path statuses Supported Not supported -- Accessing HDLM devices Supported Supported The devices must be allocated to the relevant zone.
desired kernel mode. If you switch the kernel mode from 32 bits to 64 bits when only the 32-bit version of HDLM is installed, you must reinstall HDLM after the switch is completed. For details on how to switch the kernel mode, see Switch the Kernel Mode on page 3-48 in Preparations for a New Installation of HDLM on page 3-45. • When installing HDLM on a host where a Device Manager Agent 5.
You can use zoning of Fibre Channel switches or storage-system's WWN security function to separately use HBA ports. You can also use point-to-point connections to physically separate the HBA ports. Notes on Command Execution When HDLM is used, you cannot specify the -r option in the drvconfig, devlinks, disks, or devfsadm command to reconfigure the HDLM devices on another boot disk. To reconfigure HDLM devices, you must boot them from the applicable boot disk.
you can keep the correspondence with the logical device file names once installation is complete. 4. When you execute the dlmsetconf utility for creating the HDLM driver configuration definition file, and restart the host with the reconfiguration option specified, the logical device files of the sd or ssd devices to be managed by HDLM will be deleted. Back up all pertaining to the logical device files under the /dev/dsk and /dev/rdsk directories to another directory.
# umount /mntpt ¢ Clear the automatic mount setting. The following figure shows an example of how to edit the /etc/ vfstab file. Figure 3-3 An example of how to edit the /etc/vfstab file Comment the shaded portions out by placing a hash mark (#) at the beginning of the line, or delete the line. 7. Unregister the device from the volume management software. If the device is registered in VxVM, SDS, or SVM, cancel the registration.
Set Up the /kernel/drv/sd.conf File When using an sd driver, depending on the HBA driver, you need to edit the / kernel/drv/sd.conf file to recognize the disks. For details on how to edit this file, refer to the HBA documentation. When using an ssd driver, you will not need to edit this file, since the connected disks will be automatically recognized. After editing the /kernel/drv/sd.
¢ To switch to the 64-bit kernel mode: If Solaris is running, execute the following command: # reboot -- "boot-device kernel/sparcv9/unix other-bootoptions" To boot at the ok prompt, execute the following command: boot boot-device kernel/sparcv9/unix other-boot-options Set Up the /etc/system File HDLM uses a maximum of 8 KB of the stack for the kernel thread. Add the usage of HDLM to this maximum value, specify that total value as the stack size for the kernel thread, and then install HDLM.
the data. We also recommend that you specify info as the severity level for the extended facility. This example shows how to specify the recommended settings in the /etc/syslog.conf or /etc/rsyslog.conf file: local0.info /usr/local/audlog Set Up VxVM To use HDLM and VxVM on the same host, set up VxVM as follows: 1. Make sure that the logical device file under the /dev/dsk and /dev/rdsk directories is backed up.
Set Up Solaris Cluster If you are using Sun Cluster, before installing HDLM, perform the following operations to release the device group and Quorum device settings. • • Release the disk device groups that consist of the sd or ssd devices to be managed by HDLM. For details on how to unregister an SDS or SVM diskset, see Remove SDS Settings on page 3-189 or Remove SVM Settings on page 3-190 in Preparations for HDLM Removal on page 3-185. Unregistration of a diskset deletes the data in the diskset.
f. At another node in the same cluster, switch to become the superuser, and then place the node that was stopped in step 5 into the maintenance mode. The following shows an example of executing the command: host2# /usr/cluster/bin/scconf -c -q node=host1,maintstate g. Confirm that the cluster node is in the maintenance mode. Make sure that the node that is in the maintenance node has the following status: - The node is in the Offline status.
In a Solaris 11 environment, if you cannot access the publisher of the Solaris package currently in use, HDLM installation fails. If this happens, you need to perform the procedure below to disable the publisher. If you can access the publisher normally, the following procedure is not required. 1. Execute the pkg publisher command. # pkg publisher PUBLISHER TYPE STATUS URI solaris origin online http://pkg.oracle.com/solaris/release/ 2. Disable the publisher of the package currently in use.
# zoneadm list -cv ID NAME BRAND IP 0 global solaris shared 1 non-globalzone solaris excl 2. STATUS PATH running / running /zones/non-globalzone If the status is running, shut down the non-global zone and change the status to installed. # zoneadm -z non-globalzone shutdown # zoneadm list -cv ID NAME STATUS PATH BRAND IP 0 global running / solaris shared 1 non-globalzone installed /zones/non-globalzone solaris excl 3. Detach the non-global zone. # zoneadm -z non-globalzone detach 4. Install HDLM.
sample_installhdlm.conf /any-directory/sample_installhdlm For details on the installation-information settings file and the installhdlm utility, see The installhdlm Utility for Installing HDLM on page 7-33.
In this example, the shaded portion shows information for each path of the device you are removing from HDLM management. Type in the appropriate device number, and then type quit to end the format command. 5. Insert the DVD-ROM. After a few seconds, the DVD-ROM will be automatically mounted, and the mount point of the DVD-ROM will be displayed in a list of files.
# DVD-ROM-mount-point/installux.sh -f /any-directory/ installation-information-settings-file When executing the installhdlm utility # DVD-ROM-mount-point/HDLM_Solaris/installhdlm -f/anydirectory/installation-information-settings-file If execution of the utility is completed without problems, the KAPL09211-I message is output. If the restart key is set to y (to restart the computer) in the installation-information settings file, make sure that the KAPL09211-I message has been output to the installhdlm.
# rm -rf /var/opt/DynamicLinkManager 8. Execute the following command to check whether HDLM has been installed successfully: For Solaris 8, Solaris 9, or Solaris 10 # pkginfo -l PKGINST: NAME: CATEGORY: ARCH: VERSION: BASEDIR: VENDOR: ... # DLManager DLManager Dynamic Link Manager system sparc xx.x.x.xxxx / If xx.x.x.xxxx is displayed for VERSION, then the correct version of HDLM is installed. The installed version of HDLM is shown for xx.x.x.xxxx.
Type y in response to the KAPL10242-I message displayed when you execute the dlmsetconf utility. Information pertaining to LUs to which the physical path accesses will be defined in the HDLM driver configuration definition file (/kernel/drv/ dlmfdrv.conf). Do not edit this file. 11. Specify the devices that are to be excluded from management by HDLM, if there are any. This step is unnecessary if you want HDLM to manage all applicable devices.
# /opt/DynamicLinkManager/bin/dlnkmgr view -sys HDLM Version : x.x.x-xx Service Pack Version : Load Balance : on(extended lio) Support Cluster : Elog Level : 3 Elog File Size (KB) : 9900 Number Of Elog Files : 2 Trace Level : 0 Trace File Size(KB) : 1000 Number Of Trace Files : 4 Path Health Checking : on(30) Auto Failback : on(1) Intermittent Error Monitor : off Dynamic I/O Path Control : off(10) HDLM Manager Ver WakeupTime Alive x.x.
the HDLM device) and iLU (the internal LU number). Then, compare the correspondence of the logical device file names and internal LUs that took place before HDLM was installed, and then re-register the correct HDLM devices in those programs that will use these devices. 19.
21. Based on the results of step 20, compare the logical device file names from before and after the HDLM installation, and re-register the correct HDLM devices in programs that use HDLM devices. Once installation is complete, check the path information according to the procedures listed in Checking the Path Configuration on page 3-139 to check whether HDLM has been installed correctly or not.
In this example, the shaded portion shows information for each path of the device you are removing from HDLM management. Type in the appropriate device number, and then type quit to end the format command. 5. If the settings of a Quorum device or disk device group are specified for the device to be managed by HDLM, remove those settings.
If no key is displayed, perform the steps beginning with step 11. 7. Stop all nodes that comprise the cluster. On any one node, execute the following command: # /usr/cluster/bin/scshutdown -g0 -y 8. Start all nodes comprising the cluster in non-cluster mode. Execute boot -x at the ok prompt. 9. Cancel the key registration for the device to be managed by HDLM.
cdrom/cdrom0 12. Perform the installation. ¢ If you are not performing an unattended installation of HDLM, execute either of the following installation commands: When executing the installux.sh script # DVD-ROM-mount-point/installux.sh When executing the installhdlm command # DVD-ROM-mount-point/HDLM_Solaris/installhdlm ¢ If you are performing an unattended installation of HDLM, set the installation-information settings file with the -f option specified as shown below. When executing the installux.
¢ When a correct license key is entered, the installation program displays a message that indicates the license was installed. Take care that if you enter an incorrect license key three times, installation stops. If you interrupt an installation, check whether the /var/opt/ DynamicLinkManager directory exists. If the /var/opt/ DynamicLinkManager directory exists, execute the following command to delete it: ¢ # rm -rf /var/opt/DynamicLinkManager 14.
HDLM on page 3-45 in Preparations for a New Installation of HDLM on page 3-45. 16. Execute the following dlmsetconf utility to create the HDLM driver configuration definition file (kernel/drv/dlmfdrv.conf) and the HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/config/ dlmfdrv.unconf). # /opt/DynamicLinkManager/bin/dlmsetconf Type y in response to the KAPL10242-I message displayed when you execute the dlmsetconf utility.
19. Stop all nodes that comprise the cluster. If one or more keys were displayed at step 6 and the node was started in non-cluster mode at step 8, execute the following command on all nodes: # shutdown -g0 -y -i0 If the nodes were started in cluster mode, execute the following command on any one node: # /usr/cluster/bin/scshutdown -g0 -y 20. With the reconfiguration option specified, start all nodes comprising the cluster in cluster mode, at the same time. Execute boot -r at the ok prompt.
= view, completion time = yyyy/mm/dd hh:mm:ss # Nothing is displayed for Support Cluster even if you are using cluster software. However, the cluster-support functions are, in fact, operating without problem. 22. By checking the execution results of the view operation, verify the version of the installed HDLM. If HDLM Version is x.x.x-xx, the installed version of HDLM is correct. The installed version of HDLM is shown for x.x.x-xx. 23.
Table 3-22 Information displayed by the dlmlisthdev utility for assisting HDLM transitions Item Description OldHDevName Indicates the HDLM 04-00 logical device file name (HDevName). If a hyphen (-) is displayed, the corresponding HDev did not exist before the new installation of HDLM. NewHDevName Indicates the logical device file name (HDevName) of the newly installed HDLM. If a hyphen (-) is displayed, the corresponding HDev does not exist after the new installation of HDLM.
¢ In a two-node configuration, execute the following command to release the cluster installation mode: # /usr/cluster/bin/scconf -c -q reset 32. On all nodes in the cluster, add /opt/DynamicLinkManager/bin to the PATH environment variable in the environment settings files (such as .profile) for root permissions.
host1# mount -g /dev/global/dsk/d5s0 /global/mntpt Using the HDLM Device Unconfiguration Function When Performing a New Installation of HDLM You can use the HDLM device unconfiguration function to remove Hitachi storage system devices from HDLM management on a per-LU basis. This section explains how to use the HDLM device unconfiguration function. 1. To remove an LU from HDLM management, use a text editor to edit the HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/ config/dlmfdrv.
# /opt/DynamicLinkManager/bin/dlmsetconf -u 3. 4. Restart the host with the reconfiguration option specified. Execute one of the following commands: ¢ # reboot -- -r ¢ boot -r (at the ok prompt) Execute the following command to verify that the settings were applied correctly: # /opt/DynamicLinkManager/bin/dlnkmgr view -drv Confirm that the LDEV that was removed from HDLM management is not shown.
immediately after the upgrade installation or re-installation has finished. ¢ ¢ 1. Do not interrupt the upgrade installation or re-installation. When you install HDLM while the 32-bit kernel mode is being used, only the 32-bit version of HDLM will be installed. When you install HDLM while the 64-bit kernel mode is being used, both the 32-bit and 64-bit versions of HDLM will be installed.
cdrom/cdrom0 4. Perform the installation. ¢ If you are not performing an unattended installation of HDLM, execute either of the following installation commands: When executing the installux.sh script # DVD-ROM-mount-point/installux.sh When executing the installhdlm utility # DVD-ROM-mount-point/HDLM_Solaris/installhdlm • If you are performing an unattended installation of HDLM, set the installation-information settings file with the -f option specified as shown below. When executing the installux.
¢ ¢ 6. When a correct license key is entered, the installation program displays a message that indicates the license was installed. Take care that if you enter an incorrect license key three times, installation stops. Execute the following command to check that the package is installed, by the version number in the output results: The following shows an example of executing the command: For Solaris 8, Solaris 9, or Solaris 10 # pkginfo -l PKGINST: NAME: CATEGORY: ARCH: VERSION: BASEDIR: VENDOR: ...
¢ # reboot ¢ boot (at the ok prompt) When Sun Cluster is being used: Execute the following command at one of the nodes: # /usr/cluster/bin/scshutdown -g0 -y Then, execute the following command at the ok prompt in all nodes: boot Note You must restart the host immediately after a successful upgrade installation or re-installation has been confirmed. 9. Execute the dlnkmgr command's view operation to display the status of the programs.
Extended Least Blocks. See Setting Up the HDLM Functions on page 3-140, and select the appropriate load balancing algorithm as required. ¢ Nothing is displayed for Support Cluster even if you are using cluster software. However, the cluster-support functions are, in fact, operating without problem. 10. By checking the execution results of the view operation, verify the version of the installed HDLM. If HDLM Version is x.x.x-xx, the installed version of HDLM is correct.
In this configuration example, /dev/dsk/c2t50060E8010027A82d0s2 and /dev/dsk/c2t50060E8010027A82d1s2 are sd or ssd devices. The following table shows a setting example of LDoms domain.
To install HDLM to a host in which LDoms have been set: 1. Log in to the control domain. 2. Execute the ldm stop command to stop the guest domain. For example, to stop a guest domain ldg2, execute the command as follows: # ldm stop ldg2 3. Execute the ldm unbind command to unbind the resource of the guest domain.
4. Execute the ldm rm-vdisk command to remove virtual disks from the guest domain. For example, to remove virtual disks vdisk1 and vdisk2 from the guest domain ldg2, execute the command as follows: # ldm rm-vdisk vdisk1 ldg2 # ldm rm-vdisk vdisk2 ldg2 5. Execute the ldm rm-vdsdev command to stop exporting virtual disk backends from the service domain on the I/O domain to which HDLM will be installed.
/dev/dsk/c4t50060E8010027A82d0s2 and /dev/dsk/ c4t50060E8010027A82d1s2 are HDLM device names after exporting. 9. On the control domain, execute the ldm add-vdisk command to assign the virtual disk backends to the guest domain. For example, to assign the virtual disk backends vol1 and vol2 exported from the service domain ldg1 to the guest domain ldg2, execute the command as follows: # ldm add-vdisk vdisk1 vol1@ldg1-vds0 ldg2 # ldm add-vdisk vdisk2 vol2@ldg1-vds0 ldg2 10.
Overview of Configuring a Boot Disk Environment To configure a boot disk environment in which HDLM manages the boot disk, prepare a separate disk that is not managed by HDLM, and set up an environment including Solaris, HDLM, and the applications to be used installed on that disk. If you already have such an environment, you do not need to set it up again. In this section, such an environment is called the premigration environment or pre-migration.
• Slice 0: / (root) file system • Slice 1: Swap device and dump device • Slice 3: Sun Cluster global device file system • Slice 6: /usr file system • Slice 7: Replica of the SDS or SVM state database For details on procedures involving Solaris functionality, see your Solaris documentation. Migration from an Existing HDLM Environment Perform the steps below in the pre-migration environment. The following figure shows the workflow.
1. Allocate the LUs to be used by the boot disk in the post-migration environment. 2. Get Solaris to recognize the LUs to be used by the post-migration boot disk. 3. Set the LU slices to be used by the post-migration boot disk. 4. Execute the dlmsetconf utility as follows to create the HDLM driver configuration definition file (/kernel/drv/dlmfdrv.conf) and the HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/config/ dlmfdrv.
Figure 3-9 Flow of setting up the boot disk environment when installing HDLM in the existing local boot disk environment 1. Allocate the LUs to be used by the boot disk in the post-migration environment. Configure the LU or LUs to be used by the boot disk as follows: ¢ ¢ 2. 3-86 To the slices of a single LU, allocate the Solaris file system, swap device, dump device, Sun Cluster global device file system, and replica of the SDS or SVM state database.
3. Set the LU slices to be used by the post-migration boot disk. 4. Complete the preparations for installing HDLM in the pre-migration environment and then install HDLM, referring to the description in Installing HDLM on page 3-45. Perform the necessary steps.
Figure 3-10 Flow of setting up the boot disk environment when using an existing boot disk environment without HDLM installed as the premigration environment 1. Allocate the LUs to be used by the boot disk in the post-migration environment. Configure the LU or LUs to be used by the boot disk as follows: ¢ ¢ 3-88 To the slices of a single LU, allocate the Solaris file system, swap device, dump device, Sun Cluster global device file system, and replica of the SDS or SVM state database.
2. Get Solaris to recognize the LUs to be used by the post-migration boot disk. 3. Set the LU slices to be used by the post-migration boot disk. 4. Complete the preparations for installing HDLM in the pre-migration environment and then install HDLM, referring to the description in Installing HDLM on page 3-45. Perform the necessary steps.
Figure 3-11 Flow of setting up the boot disk environment when building a new pre-migration environment 3-90 1. Install Solaris in the pre-migration environment. 2. Install Sun Cluster if required. For details, see your Sun Cluster documentation. 3. Install VxVM if required. For details, see your VxVM documentation. 4. Allocate the LUs to be used by the boot disk in the post-migration environment.
¢ ¢ To the slices of a single LU, allocate the Solaris file system, swap devices, dump devices, Sun Cluster global device file system, and replica of the SDS or SVM state database. To the slices of multiple LUs, allocate the Solaris file system, swap devices, dump devices, Sun Cluster global device file system, and replica of the SDS or SVM state database. In this case, make sure that the physical path of each LU is the same. 5.
# newfs /dev/rdsk/c8t50060E80004361D1d1s3 3. Mount the Solaris file system in the post-migration environment. If using Sun Cluster, also mount the Sun Cluster global device file system. The following shows an example of executing the command: # mount /dev/dsk/c8t50060E80004361D1d1s0 /mnt0 # mount /dev/dsk/c8t50060E80004361D1d1s6 /mnt1 # mount /dev/dsk/c8t50060E80004361D1d1s3 /mnt2 4. Copy the Solaris file system from the pre-migration environment to the post-migration environment.
/dev/dsk/c0t0d0s0 /dev/rdsk/c0t0d0s0 / ufs 1 no /dev/dsk/c0t0d0s6 /dev/rdsk/c0t0d0s6 /usr ufs 1 no /dev/did/dsk/d24s3 /dev/did/rdsk/d24s3 /global/.devices/ node@1 ufs 2 no global After editing /dev/dsk/c8t50060E80004361D1d1s0 /dev/rdsk/ c8t50060E80004361D1d1s0 / ufs 1 no /dev/dsk/c8t50060E80004361D1d1s6 /dev/rdsk/ c8t50060E80004361D1d1s6 /usr ufs 1 no /dev/did/dsk/d30s3 /dev/did/rdsk/d30s3 /global/.
As the boot device, you can specify any owner path of the boot disk. ¢ In Solaris 10 or Solaris 11: As the boot device, you can only specify a boot disk path used by HDLM to determine the logical device file name of the HDLM device. To find this path, perform the procedure described in When the Path Configuration Is Changed on page 4-17 in Precautions Regarding Changes to the Configuration of an HDLM Operating Environment on page 4-17.
Note In Solaris 8 or Solaris 9, find and make a note of the boot device names to specify in the OBP for all owner paths in case an error occurs on the path specified for the boot device. 12. Boot the system from the post-migration environment. ¢ ¢ ¢ Perform steps 13 to 17 only if SDS or SVM is being used and you are migrating a replica of the SDS or SVM state database to the postmigration environment. In this case, keep the pre-migration environment recognized by Solaris.
18. Back up the contents of both the pre-migration and post-migration environments. Perform a backup so that you can take action in the following situations: ¢ ¢ Recover when an error occurs in the post-migration environment. Boot from the pre-migration environment, and repair the postmigration environment. The post-migration backup can be used to restore the post-migration environment.
For Solaris 10 Figure 3-12 Creating a ZFS boot disk environment 1. Copy the local boot disk environment from the ZFS file system to the LUs (SCSI device) in the storage system. 2. Replace the LUs on the SCSI device in the storage system with those on the HDLM device. You can also use the above procedure to create a boot disk when the premigration boot disk is in the storage system rather than on the local disk, or when the file system is a UFS file system.
For Solaris 11 Figure 3-13 Creating a ZFS boot disk environment 1. Migrate the local boot disk environment in ZFS to an LU (HDLM device) in the storage system. 2. Configure the ZFS boot disk environment after the migration. Creating a ZFS Boot Disk Environment (for Solaris 10) The following describes the procedure for creating a ZFS boot disk environment for Solaris 10.
Copying the local boot disk environment to the LUs (SCSI device) in the storage system Copy the local boot disk environment to the LUs (SCSI device) in the storage system. The following describes the preparations required before you can start copying. Preparations before copying: • Two LUs have the same capacity as the local boot disk need to be created in the storage system. You also need to configure at least one data disk on the host.
In the following example, the boot disk environment name is zfsBE, and the root pool name is rpool2. # lucreate -n zfsBE -p rpool2 When the command terminates normally, the boot disk has been copied to rpool2. 5. Execute the lustatus command to check the copied ZFS boot disk environment. # lustatus Boot Environment Is Active Active Can Copy Name Complete Now On Reboot Delete Status --------------------------------------------------------------s10s_u9wos_14a yes yes yes no zfsBE yes no no yes - 6.
migration root pool. In addition, the pre-migration UFS file system remains mounted in the post-migration environment. If you do not want to use the pre-migration UFS file system, copy the data in the file system to the post-migration root pool. If you do not know how to copy the data of the UFS file system or how to unmount the premigration UFS file system, contact the company with which you have a contract for Oracle Support Services.
3. Execute the zpool status command to confirm that the replacement has been completed. # zpool status When the replacement has been completed, the following message is output: scrub: resilver completed after hh:mm with 0 errors hh:mm is the time that has passed since the replacement was completed. If the above message is not output, wait a while, and then re-execute the command to confirm that the replacement has been completed. 4.
# zpool create hdlmpool /dev/dsk/c6t50060E8010027A80d7s0 2. Execute the zpool status command and make sure that the root pool has been created correctly. # zpool pool: state: scan: config: status hdlmpool hdlmpool ONLINE none requested NAME hdlmpool c6t50060E8010027A80d7s0 3. STATE ONLINE ONLINE READ WRITE CKSUM 0 0 0 0 0 0 Execute the installboot command to copy the boot block so that you can boot the OS from the ZFS boot disk environment.
# eeprom boot-device=/pci@7c0/pci@0/pci@1/pci@0,2/SUNW,qlc@2,1/ fp@0,0/disk@w50060e8010027a80,7:a For how to set a boot device name, see step 11 in Setting Up the PostMigration Environment on page 3-91 of Procedure for Configuring a Boot Disk Environment on page 3-83. For how to obtain a device name to be set for OBP, see the documentation of the HBA you are using. 8. Execute the init command and restart the host. # init 6 Do not use the reboot command to restart the host. 9.
Before editing /dev/zvol/dsk/rpool/swap swap - no - no - After editing /dev/zvol/dsk/hdlmpool/swap swap - 3. Execute the zfs create command to create a dump device for a file system in the root pool (hdlmpool) after the migration. The dump device size varies depending on the operating environment. # zfs create -V 3G hdlmpool/dump 4. Execute the dumpadm command to switch to the dump device in the ZFS boot disk environment after the migration. # dumpadm -d /dev/zvol/dsk/hdlmpool/dump 5.
3. Execute the zfs list command to make sure the data is copied to the root pool (hdlmpool) in the ZFS boot disk environment after the migration. # zfs list -r hdlmpool NAME hdlmpool hdlmpool/ROOT hdlmpool/ROOT/hdlmboot hdlmpool/ROOT/hdlmboot/var hdlmpool/VARSHARE hdlmpool/dump hdlmpool/export hdlmpool/export/home hdlmpool/swap USED 12.5G 5.27G 5.27G 1.04G 43K 3.09G 67K 35K 4.13G AVAIL 12.0G 12.0G 12.0G 12.0G 12.0G 12.1G 12.0G 12.0G 12.1G REFER 73.5K 31K 4.20G 1.04G 43K 3.00G 32K 35K 4.
c6t50060E8010027A80d8s0 -> ../../devices/pseudo/dlmndrv@1/dlmfdrv@w50060e8010027a80,8:a 2. Execute the zpool replace command to change the current ZFS boot disk environment from the HDLM logical device specification to the HDLM physical device specification. # zpool replace root-pool-name current-boot-disk-environmentdevice HDLM-physical-device The physical device file of the HDLM device is the device file in the / devices directory.
The following is an example when the post-replacement HDLM device is c6t50060E8010027A80d8s0: # installboot -F zfs /usr/platform/`uname -i`/lib/fs/zfs/bootblk /dev/rdsk/c6t50060E8010027A80d8s0 5. Execute the eeprom command to specify the post-replacement device as the boot device.
PathID PathName DskName iLU ChaPort Status Type IO-Count IO-Errors DNum HDevName 000000 0000.0000.0000000000641000.000B HITACHI .DF600F 85011299 0008 0A Online Own 8840 0 - c6t50060E8010027A80d8 000001 0000.0000.0000000000641400.000B HITACHI .DF600F 85011299 0008 1A Online Own 1506 0 - c6t50060E8010027A80d8 ... # . . Make sure the number of I/Os are counted for the HDLM device (the underlined part of c6t50060E8010027A80d8) specified as the boot device.
In the following example, the boot disk environment name is zfsBE2 and the name of the device on the built-in disk is c1t0d0s0: # luactivate zfsBE2 # installboot -F zfs /usr/platform/`uname -i`/lib/fs/zfs/ bootblk /dev/rdsk/c1t0d0s0 5. Execute the init command, and then restart the host. # init 6 Do not use the reboot command to restart the host. 6. Execute the zpool export command to export the root pool from the pre-migration local boot disk environment.
2. Execute the zpool status command to check whether replacement is complete. # zpool status hdlmpool pool: hdlmpool state: ONLINE scan: resilvered 12.3G in 0h2m with 0 errors on Fri Dec 13 16:35:25 2013 config: NAME hdlmpool c6t50060E8010027A80d7s0 STATE ONLINE ONLINE READ WRITE CKSUM 0 0 0 0 0 0 errors: No known data errors If replacement is complete, the following message appears: scan: resilvered XXX in XhYm with 0 errors XhYm indicates the amount of time that elapsed before replacement completion.
Creating a New Boot Environment 1. Execute the zpool import command to check the ID of the root pool (rpool). # zpool import pool: rpool id: 6462316201492788659 state: ONLINE status: The pool is formatted using an older on-disk version. action: The pool can be imported using its name or numeric identifier, though some features will not be available without an explicit 'zpool upgrade'. config: rpool c1t0d0s0 2.
# beadm create -p rpool -a solaris-2 5. Execute the beadm list command to check that the ZFS boot disk environment was created. Make sure that R appears in the Active column of the hdlmboot row for the created ZFS boot disk environment. # beadm list BE Active Mountpoint Space Policy Created ------- ---------- ----- ------ ------hdlmboot NR / 2.06G static 2011-12-06 11:37 solaris 2.93G static 2011-11-29 14:42 solaris-2 R 4.84G static 2011-12-11 14:01 6.
disk in the pre-migration environment to the ZFS boot disk in the postmigration environment. The following is an example: Before editing /dev/zvol/dsk/hdlmpool/swap - - swap - no After editing /dev/zvol/dsk/rpool/swap - - swap - no 3. If rpool/dump does not exist as a file system, execute the zfs create command to create a dump device for the file system in the postmigration root pool rpool. The dump device size differs depending on the operating environment. # zfs create -V 3G rpool/dump 4.
# zfs destroy –r rpool/export # zfs send -vRp hdlmpool/export@2011-12-07.01 | zfs receive uvdF rpool 3. Execute the zfs list command to check that data was successfully copied to rpool, which is the root pool of the post-migration ZFS boot disk environment. # zfs list -r rpool NAME rpool rpool/ROOT rpool/ROOT/solaris rpool/ROOT/solaris-2 rpool/ROOT/solaris-2/var rpool/ROOT/solaris/var rpool/VARSHARE rpool/dump rpool/export rpool/export/home rpool/swap USED 42.7G 21.9G 21.5M 102M 1K 18.2M 216M 16.
# df files / (rpool/ROOT/rboot):124234885 blocks 124234885 files /var (rpool/ROOT/rboot/var):124234885 blocks 124234885 /export (rpool/export ):124234885 blocks 124234885 files /export/home (rpool/export/home):124234885 blocks 124234885 files /rpool (rpool ):124234885 blocks 124234885 files Make sure the / (root) directory, the /var directory, the /export directory, and the /export/home directory are mounted in the root pool (hdlmpool) after the migration. 2.
Figure 3-14 Migration flow when coping the file system from a boot disk environment to the local boot disk environment 1. Boot the system from the pre-migration environment. Perform steps 1 to 16 in the pre-migration environment. 2. Allocate the LUs to be used by the boot disk in the post-migration environment. 3. Get Solaris to recognize the LUs to be used by the post-migration boot disk. 4. Set the LU slices to be used by the post-migration boot disk. ¢ If using Sun Cluster, go to the next step.
5. Assign device IDs. The following shows an example of executing the command: # /usr/cluster/bin/scdidadm -r 6. Notify the server of the new device ID information. The following shows an example of executing the command: # /usr/cluster/bin/scdidadm -ui 7. Update the Sun Cluster global device namespace. The following shows an example of executing the command: # /usr/cluster/bin/scgdevs 8.
rf -) 13. In the post-migration /etc/system file, delete the rootdev parameter that specifies the HDLM physical device file name. 14. Edit the post-migration boot disk's /etc/vfstab file, changing the Solaris file system to the HDLM logical device file name of the post-migration boot disk. If using Sun Cluster, change the mount device# of the Sun Cluster global device file system in the same way. An example is shown below.
Before editing DUMPADM_DEVICE=/dev/dsk/c8t50060E80004361D1d1s1 After editing DUMPADM_DEVICE=/dev/dsk/c0t0d0s1 17. Boot the system from the post-migration environment. ¢ ¢ ¢ Skip to step 18 if SDS or SVM is being used and you are migrating a replica of the SDS or SVM state database to the post-migration environment. In this case, keep the pre-migration environment recognized by Solaris. Skip to step 21 if SDS or SVM is being used but you are not migrating a replica of the SDS or SVM state database.
¢ ¢ VxVM Sun StorageTek Traffic Manager software (or, in Solaris 10, the Solaris multipathing software MPxIO) Configuring a Mirrored Boot Disk Environment Incorporating SVM A mirrored boot disk incorporating SVM can be managed by HDLM. The following configuration is supported: • OS: Solaris 10 • RAID level: Mirroring (no more than three mirrors) • Cluster: None Precautions • When you use the dlmsetconf utility, do not specify the -r option.
• a. Remove HDLM after migration to a local boot disk environment by performing the procedures in Removing HDLM from the Boot Disk Environment on page 3-193 in Removing HDLM on page 3-191. b. Upgrade Solaris. c. Configure a boot disk environment again by performing the procedure in Configuring a Boot Disk Environment in Which HDLM Manages the Boot Disk and Mirroring the Environment by Using SVM on page 3-122.
Figure 3-15 Configuring a boot disk environment in which HDLM manages the boot disk, from the local boot disk environment Mirroring a Boot Disk Environment in Which HDLM Manages the Boot Disk by Using SVM Mirror a configured boot disk by using SVM. The following figure provides an overview of mirroring a boot disk environment.
• Slice 0: / (root) file system • Slice 1: Swap device and dump device • Slice 6: /usr file system • Slice 7: SVM state database 1. Prepare another LU with the same capacity as that of the created boot disk, and make the configuration and capacity of slices the same. 2. Create the SVM state database. # matadb -a -f c8t50060E08000436D1d4s7 # metadb -a c8t50060E08000436D1d5s7 3. Decide the metadevice names. In this procedure, specify the settings as follows: ¢ 4.
7. Execute the metaroot command. # metaroot d10 8. Reboot the host. 9. When the host starts, check the swap device and dump device. Then execute the mount command to confirm that / (root) and /usr have been correctly configured.
Stripe 0: Device Dbase State Reloc Hot Spare c8t50060E08000436D1d4s0 0 d12: Submirror of d10 State: Okay Size: 25190400 blocks (12 GB) Stripe 0: Device Dbase State Reloc Hot Spare c8t50060E08000436D1d5s0 0 ... Start Block /dev/dsk/ No Okay Yes Start Block /dev/dsk/ No Okay Yes 12. Execute the dlnkmgr command's view -path operation to confirm that an I/O is issued to the LU set as the boot disk.
Installing HDLM and then Configuring the Environment 1. Install HDLM by referring to Installing HDLM on page 3-45. 2. Execute the dlmsetconf utility for creating the HDLM driver configuration definition file to configure the HDLM device. 3. Remove the LUs of the mirrored boot disks from HDLM management. For details, see Using the HDLM Device Unconfiguration Function When Performing a New Installation of HDLM on page 3-72. 4. Reboot the host with the reconfiguration option specified.
• Slice 7: SVM state database 1. Decide the metadevice names. In this section, it is assumed that the metadevice names of the existing boot disks and the boot disks after the migration are as shown in following table. Table 3-24 Metadevice names Boot disk Current File system /(root) Mirror d10 Sub mirror d11 d12 swap d20 d21 d22 /usr d30 d31 d32 After migration /(root) d100 d101 d102 swap d200 d201 d202 /usr d300 d301 d302 2.
# newfs /dev/md/rdsk/d300 5. Mount the Solaris basic file system of the post-migration environment. For example: # mount /dev/md/dsk/d100 /mnt0 # mount /dev/md/dsk/d300 /mnt1 6. Copy the existing Solaris basic file system to the post-migration environment. For example: # ufsdump 0f - / | (cd /mnt0; ufsrestore rf -) # ufsdump 0f - /usr | (cd /mnt1; ufsrestore rf -) 7.
To determine this path, perform the procedure in When the Path Configuration Is Changed on page 4-17 in Precautions Regarding Changes to the Configuration of an HDLM Operating Environment on page 4-17. To obtain the boot device name specified by OBP, edit the character string of the determined path as follows: ¢ ¢ Replace the driver name (ssd) of an ssd driver with disk. Add the slice identifier of the / (root) file system (:a value from :a to :h) to the end.
# swap -l swapfile /dev/md/dsk/d200 # dumpadm Dump content: Dump device: Savecore directory: Savecore enabled: dev swaplo blocks free 85,200 16 8409584 8409584 kernel pages /dev/md/dsk/d200 (swap) /var/crash/host yes # mount / on /dev/md/dsk/d100 read/write/setuid/devices/intr/largefiles/ logging/xattr/onerror=panic/dev=1540064 on Mon Mar 9 11:00:25 2009 /usr on /dev/md/dsk/d300 read/write/setuid/devices/intr/ largefiles/logging/xattr/onerror=panic/dev=154012c on Mon Mar 9 11:00:25 2009 13.
Size: 25190400 blocks (12 GB) d101: Submirror of d100 State: Okay Size: 25190400 blocks (12 GB) Stripe 0: Device Dbase State Reloc Hot Spare /dev/dsk/c8t50060E08000436D1d6s0 No Okay Yes d102: Submirror of d100 State: Okay Size: 25190400 blocks (12 GB) Stripe 0: Device Dbase State Reloc Hot Spare /dev/dsk/c8t50060E08000436D1d7s0 No Okay Yes ... Start Block 0 Start Block 0 17. Execute the dlnkmgr command's view -path operation to confirm that I/O is issued to the LU set as the boot disk.
Backups of the environments are required for the following situations: ¢ ¢ If an error occurs in the post-migration environment: Reboot from the pre-migration environment, and then use it to restore the post-migration environment. Alternatively, the backup of the post-migration environment can also be restored. If you want to migrate from a boot disk environment to a local boot disk environment, you can use the backup to return to the state before creating the boot disk environment.
Figure 3-19 Current boot disks and boot disks after the migration The LU device names of the current boot disks are c8t50060E08000436D1d4 and c8t50060E08000436D1d5, and the LU device names of the boot disk after the migration are c6t50060E08000436D1d6 and c6t50060E08000436D1d7, as shown in the above figure.
Boot disk File system /usr Mirror d300 Sub mirror d301 d302 After migration /(root) d10 d11 d12 swap d20 d21 d22 /usr d30 d31 d32 2. Create metadevices on the sd/ssd device, and then mirror them. (Single mirror) # metainit -f d11 1 1 c6t50060E08000436D1d6s0 # metainit d10 -m d11 # metainit -f d21 1 1 c6t50060E08000436D1d6s1 # metainit d20 -m d21 # metainit -f d31 1 1 c6t50060E08000436D1d6s6 # metainit d30 -m d31 3.
# ufsdump 0f - /usr | (cd /mnt1; ufsrestore rf -) 7. Edit the file /etc/vfstab (/mnt0/etc/vfstab) of the boot disk after the migration to change the definition of the swap after the migration to a metadevice (/dev/md/dsk/d20) and change the definition of /usr after the migration to a metadevice (/dev/md/ds/d30). Before change: /dev/md/dsk/d200 swap no /dev/md/dsk/d300 /dev/md/rdsk/d30 /usr ufs 1 - no - After change: /dev/md/dsk/d20 swap no /dev/md/dsk/d30 /dev/md/rdsk/d300 /usr ufs 1 8.
Figure 3-20 Booting from the post-migration environment 11. When the host starts, confirm that the swap device and dump device have been correctly configured. Then, use the mount command to confirm that / (root) and /usr have been correctly configured.
# metainit -f d22 1 1 c6t50060E08000436D1d7s1 # metattach d20 d22 # metainit -f d32 1 1 c6t50060E08000436D1d7s6 # metattach d30 d32 13. Copy the SVM state database to the boot disk after migration.. # metadb -a c6t50060E08000436D1d6s7 # metadb -a c6t50060E08000436D1d7s7 14. Delete the SVM in the pre-migration environment. # metaclear d100 d101 d102 # metaclear d200 d201 d202 # metaclear d300 d301 d302 # metadb -d c8t50060E08000436D1d4s7 # metadb -d c8t50060E08000436D1d5s7 15.
Remove HDLM by referring to Removing HDLM on page 3-184. Checking the Path Configuration HDLM functions, such as load balancing and failover, are only available for HDLM management-target devices that have more than one active path. After you install HDLM or change the hardware configuration, check the structure and statuses of the paths. To check the path information, use the HDLM command's view operation. The following describes how to check path information by using the dlnkmgr command's view operation.
Setting Up HDLM Functions HDLM includes functions like the load balancing function, the automatic failback function, the error logging function, the audit logging function, etc. You can set up these functions by using the dlnkmgr command's set operation. The following subsections describe these setup methods.
Table 3-26 The default and recommended values for each function Function Load-balancing Default value Recommended value on on The Extended Least I/Os algorithm is used. The recommended algorithm depends on the operating environment. on on 30-minute check interval The recommended checking interval depends on the operating environment. on 1-minute check interval The recommended checking interval depends on the operating environment.
The following is an example command to set load-balancing: # /opt/DynamicLinkManager/bin/dlnkmgr set -lb on -lbtype exlio When load balancing is enabled by specifying on for the -lb parameter, specify one of the following algorithm values after the -lbtype parameter: • rr for the Round Robin algorithm • exrr for the Extended Round Robin algorithm • lio for the Least I/Os algorithm • exlio for the Extended Least I/Os algorithm • lbk for the Least Blocks algorithm • exlbk for the Extended Least Bl
If this condition is not satisfied, an error will occur and the warning message KAPL01080-W will be output. If this happens, change any of the following settings: the checking interval for automatic failbacks, the intermittent error-monitoring interval, or the number of times that the error needs to occur. If you set the number of times that the error needs to occur to 1, the above condition does not need to be satisfied.
To enable intermittent error monitoring, set it to on. Otherwise, set it to off. When on is specified, you can use the -intvl and -iemnum parameters to specify intermittent error conditions (the conditions used by the system to determine whether an intermittent error is occurring). Specify the monitoring interval for an intermittent error in the -intvl parameter, and the number of times that the error needs to occur in the -iemnum parameter.
The higher this value is set, the more information that will be output. As the amount of log information to be output increases, it will take less time to overwrite the old error log information with the new information. The following is an example of setting up the error log collection level by using a command: # /opt/DynamicLinkManager/bin/dlnkmgr set -ellv 2 Specify the error log collection level as a number. Setting the Trace Level You can set up the trace level for a trace file hdlmtrn.
You can specify a value (in kilobytes) from 100 to 2000000 for the error log file size. The specified value is applied for HDLM manager logs. If each error log file reaches the specified size, the information in the old error log file will be wrapped around and will be overwritten with new information. By specifying both the log file size and the number of log files, you can collect up to 32000000 kilobytes (approximately 30GB) of error logs in total.
# /opt/DynamicLinkManager/bin/dlnkmgr set -systfs 2000 Specify the size of the trace file in kilobytes. Setting the Number of Trace Files Trace files for which you can set the number of files are hdlmtrn.log (n indicates a file number from 1 to 64). For the number of the trace files, you can specify a value from 2 to 64.
Table 3-30 Values Indicating Audit Log Data Categories Value Explanation ss Audit log events of the StartStop category are collected. a Audit log events of the Authentication category are collected. ca Audit log events of the ConfigurationAccess category are collected. all Audit log events of the StartStop, Authentication, and ConfigurationAccess categories are all collected.
Value local7 or 23 Corresponding facility value in the /etc/syslog.conf file local7 This example shows how to specify the audit log facility: # /opt/DynamicLinkManager/bin/dlnkmgr set -audfac local0 Checking the Updated Settings This chapter describes how to check the updated settings by using the dlnkmgr command's set operation after settings have been changed. When you change some settings, you can display information about all of HDLM function settings.
(HNTRLib2) (/var/opt/hitachi/HNTRLib2/spool/hntr2n.log (n indicates a file number)). If a lot of integrated trace information is output, the older information might end up getting deleted in a very short amount of time. Also, if a large amount of integrated trace information is suddenly all output at the same time, any integrated trace information that is overflowing the buffer might not be saved into the integrated trace files.
Displaying the Hitachi Network Objectplaza Trace Library setup Menu To display the Hitachi Network Objectplaza Trace Library setup menu: 1. Log in as a user with root permissions. 2. Execute the following command: # /opt/hitachi/HNTRLib2/bin/hntr2utl2 The Hitachi Network Objectplaza Trace Library setup menu appears. If you do not want to change the settings, type e and then press the Enter key to quit the menu. The following explains how to modify each setting.
the Buffer Size Per Monitoring Interval Duration on page 3-152. The recommend value is 4096 for collecting an integrated trace. If you do not want to change the integrated trace file size, leave New Size(KB) blank, type !, and then press the Enter key to return to the Hitachi Network Objectplaza Trace Library setup menu. 3. Press the Enter key. The new setting is applied and the Hitachi Network Objectplaza Trace Library setup menu appears again.
2. Enter a larger size in New Size(KB). Set a new buffer size to fit the monitoring interval set in 5: Interval Timer. The default value is 64. You can specify a value from 8 to 2048 (KB). Set this to a value smaller than the integrated trace file size set in step 2 of Changing the Size of Integrated Trace Files on page 3-151. The recommended value is 256 for collecting an integrated trace.
A screen to set the monitoring interval for the number of messages output to the integrated trace file appears. The current value is displayed in Current Span (sec). 3-154 2. Enter a desired interval in New Span(sec). You can specify a value from 0 to 3600 (seconds). The default and recommended value is 0. If you do not want to change the monitoring interval, leave New Span(sec) blank, type !, and then press the Enter key to return to the Hitachi Network Objectplaza Trace Library setup menu.
If you do not want to change the maximum number of messages output, leave New Max(sec) blank, enter !, and then press the Enter key to return to the Hitachi Network Objectplaza Trace Library setup menu. 6. Press the Enter key. The new setting is applied and the Hitachi Network Objectplaza Trace Library setup menu appears again. Finishing the Hitachi Network Objectplaza Trace Library Settings To close the Hitachi Network Objectplaza Trace Library setup menu when you are finished: 1.
# /opt/hitachi/HNTRLib2/bin/hntr2mon -d & 7. Start the programs stopped in step 3. If you stopped programs other than HDLM in step 3, start them. Creating File Systems for HDLM (When Volume Management Software Is Not Used) When volume management software is not used, the file system directly uses HDLM devices. This section explains how to mount an HDLM-managed device as a UFS file system by specifying the name of the logical device file for an HDLM device.
Figure 3-21 An output example for the mount -p command Edit the /etc/vfstab file as appropriate based on the settings output. The shaded portion indicates the part of the /etc/vfstab file to set. The following figure shows an example of how to edit the /etc/vfstab file. Figure 3-22 An example of how to edit the /etc/vfstab file Add the shaded portion to the file. Setting Up VxVM When you are using VxVM, VxVM accesses an LU by using an HDLM device. VxVM will automatically recognize HDLM devices.
To register an HDLM device: 1. Make sure that VxVM recognizes HDLM devices. Execute the following commands sequentially to check whether VxVM recognizes HDLM devices. ¢ The dlnkmgr command's view -path operation Check the HDLM devices that must be recognized by VxVM. The following shows an example of executing the command: The following describes the procedure for checking whether the shaded HDLM devices are recognized. ¢ vxdisk command Make sure that the devices are recognized by VxVM.
Note The device name of the enclosure corresponding to the LU might vary before and after installing HDLM. Make sure that you perform step 1 to check the correspondence between the device name of the enclosure and the LU. If the vxdisk command displays the value of HDevName (cUtXdY) output by the dlnkmgr command's view -path operation, the HDLM device is recognized by VxVM. Go to step 2.
Note Do not add an HDLM device to a disk group registered as bootdg. Also, do not register a disk group containing an HDLM device in bootdg. 4. If you are using Sun Cluster, register the created disk group into a Sun Cluster disk device group. For details, see Linking VxVM and Solaris Cluster on page 3-174. Creating VxVM Volumes A VxVM volume can be created by using the vxassist (or vxmake) command or Veritas Enterprise Administrator. For details, refer to the VxVM documentation.
Devices to Be Removed from VxVM The following figure shows devices that are recognized by VxVM before they are disabled. Figure 3-23 Devices That Are Recognized by VxVM before They Are Removed from VxVM You must remove the following non-HDLM devices that are recognized by VxVM. If the following devices are being managed by VxVM, HDLM functions, such as load balancing and failover, will not work normally.
# For logical device files of HDLM devices, the number of such files to be created is the same as the number of slices for one LU. The following figure shows the devices that are recognized by VxVM after the non-HDLM devices are disabled. Figure 3-24 Devices Recognized by VxVM after Removal of Devices from VxVM Only HDLM devices (c5t0d0) that are actually created are recognized. In this case, HDLM functions, such as load balancing and failover, work normally.
the configuration example shown in Figure 3-25 When HBA ports on the alternate path differ from those on the primary path connected to the Hitachi storage system on page 3-163. This figure shows a configuration that is recognized by VxVM before devices are removed. Figure 3-25 When HBA ports on the alternate path differ from those on the primary path connected to the Hitachi storage system To remove devices from VxVM, in the above configuration: 1.
The shaded portion, in the next step, is contained in the /etc/vx/ vxvm.exclude file. 2. Make an entry for the controller of the device you wish to remove from VxVM in the controllers section of the /etc/vx/vxvm.exclude file. In the /etc/vx/vxvm.exclude file, write the shaded portion of the execution result of the dlmvxexclude utility shown in step 1. The settings will take effect when the host is restarted. The following figure shows an example of the /etc/vx/vxvm.exclude file.
The devices displayed for HDevName (the shaded portion) are the HDLM devices that must be recognized by VxVM. In the above example, two devices, c5t0d0 and c5t0d1, must be recognized. 5. Execute the vxdisk command to list devices that are recognized by VxVM. The following shows an example of executing the command: If the HDLM devices output by the dlnkmgr command's view -path operation (the shaded portion) are displayed, the HDLM devices have been recognized by VxVM.
If any non-HDLM devices are ENABLED, perform the procedure described in Actions To Be Taken if an sd or ssd Device Has Not Been Suppressed from VxVM on page 3-170 of Removing Devices from VxVM on page 3-160 to set up vxvm.exclude again. When the display resulting from the vxdmpadm command is as shown in the following table, you do not need to reconfigure vxvm.exclude. Table 3-33 When setup of vxvm.exclude is unnecessary Display of the vxdmpadm command HDLM devices are not displayed.
Figure 3-27 When the HBA Port of the Physical Path Connected to a Hitachi Storage System Is the Same As the One Connected to a Third-party Storage System To remove devices from VxVM, in the above configuration: 1. Use the dlmvxexclude utility to acquire the mapping between controller numbers and devices. The devices to be removed from VxVM are output to the paths section in the execution result of the dlmvxexclude utility.
The shaded portion indicates the devices that need to be removed from VxVM. 2. Make an entry for the device you wish to removed from VxVM in the paths section of the /etc/vx/vxvm.exclude file. In the /etc/vx/vxvm.exclude file, write the shaded portion of the execution result of the dlmvxexclude utility shown in step 1. The settings will take effect when the host is restarted. The following figure shows an example of the /etc/vx/vxvm.exclude file. Figure 3-28 Example of a description in the /etc/vx/vxvm.
The devices displayed for HDevName (the shaded portion) are the HDLM devices that must be recognized by VxVM. In the above example, it is c4t0d0. 5. Execute the vxdisk command to list devices that are recognized by VxVM. The following shows an example of executing the command: If the HDLM devices output by the dlnkmgr command's view -path operation (the shaded portion) are displayed, the HDLM devices have been recognized by VxVM.
from VxVM on page 3-170 of Removing Devices from VxVM on page 3-160 to set up vxvm.exclude again. When the display resulting from the vxdmpadm command is as shown in the following table, you do not need to reconfigure vxvm.exclude. Table 3-34 When setup of vxvm.exclude is unnecessary Display of the vxdmpadm command HDLM devices are not displayed. Why suppression is unnecessary They are non-HDLM-managed devices such as a built-in disk. Only HDLM devices are displayed.
Example 1: The vxdisk command does not display HDLM devices but displays the sd or ssd devices corresponding to those HDLM devices. The vxdisk command displays the SCSI device c3t0d0, which corresponds to an HDLM device, instead of the HDLM device c5t0d0. Example 2: The vxdmpadm command outputs a non-HDLM device, and the STATE column shows ENABLED.
The vxdmpadm command displays c3t0d0 and c4t1d0 in addition to the HDLM device c5t0d0, and the STATE column shows ENABLED. If sd or ssd devices have not been removed normally from VxVM, as shown in Examples 1 and 2, take the following action: To suppress sd or ssd devices: 1. Execute the vxdmpadm command. In the following example, HDLM device c5t0d0 has been created and a non-HDLM device is ENABLED.
Figure 3-29 The /etc/vx/vxvm.exclude file (before editing) Figure 3-30 The /etc/vx/vxvm.exclude file (after editing) Among the non-HDLM devices in the ENABLED state, there might be two or more devices whose tXdY portions in the logical device file name (cUtXdYs2 format) are the same. In this case, change the currently assigned controller number in /etc/vx/vxvm.exclude. 5. After modifying all the sd or ssd device entries that have not been suppressed in /etc/vx/vxvm.
If the STATE column shows DISABLED, the device has been removed from VxVM. The following shows an example of executing the command: If there is more than one candidate for the controller number to be assigned in step 4, you might not be able to remove devices from VxVM in one operation. In this case, change the controller number you changed in step 4 to another controller number that was displayed using the vxdmpadm command, and then perform steps 5 and 6 again.
4. Verify that the VxVM disk group has been imported into only one of the nodes that constitute a cluster. If the VxVM disk group has been deported from all nodes, import the VxVM disk group to be registered in Sun Cluster into only one of the nodes. 5. If the VxVM disk group is not registered in Sun Cluster, execute the scsetup utility from the node into which you imported the VxVM disk group to be registered in Sun Cluster, and then register the VxVM disk group into a Sun Cluster disk device group.
KAPL05023-E Could not process the IOCTL(00004D03). (0000000D:00000000) 6. If the VxVM volume has not been created, create a VxVM volume in the disk group from the master node (the node that currently acts as the master for the disk device group). The following shows an example of executing the command: # vxassist -g dlmdg make vol-dg 10g 7.
• Even if you use a local diskset or shared diskset, do not specify the -r option when changing the path configuration and executing the dlmsetconf utility for creating the HDLM driver configuration definition file. When you specify the -r option, the minor number of the HDLM device used on the SDS will be changed, and you will no longer be able to use the metadevice. • Even if you use a local diskset or shared diskset, you cannot use the DiskSuite tool (metatool). Use the command line utility instead.
host1# metadb -a -f c1t1d0s1 c1t1d0s3 c1t1d0s7 2. Create a local metadevice. Use the HDLM device (slice) to create a local metadevice (logical disk). The types of metadevices include coupled, stripe, mirrored, RAID5, and transmeta devices. The following shows an example of executing the command to create local metadevice (d0) of type stripe (1) comprised of two HDLM devices (slices). host1# metainit d0 1 2 c5t1d0s7 c5t2d1s7 3. Create a file system on the local metadevice.
host1# metaset -s diskset1 -a -h host1 host2 3. Register a disk into the shared diskset. The following shows examples of executing the commands: When not using Sun Cluster: host1# metaset -s diskset1 -a c5t1d2 c5t1d3 When using Sun Cluster: host1# metaset -s diskset1 -a /dev/did/dsk/d10 /dev/did/dsk/ d11 4. Create a shared metadevice in the shared diskset.
Setting Up SVM If SVM is installed on an HDLM host, SVM uses the logical device file name of the HDLM device to access the HDLM-managed device. Therefore, after installing HDLM, you need to register the logical device file name of the HDLM device in SVM. When removing HDLM, unregister the logical device file name of the HDLM device and set the logical device file name of the sd or ssd device. Notes When using SVM, keep the following in mind.
Registering HDLM Devices If SVM is used, perform the following settings after installing HDLM, and register the SVM volume. To Use a Local Volume To use a local volume: 1. Create a replica of the state database. Create three replicas of the local state database on the disk (slice). Create these replicas on a disk that is not managed by HDLM, not shared, and has not been connected to a Fibre Channel. The following shows an example of executing the command: host1# metadb -a -f c1t1d0s1 c1t1d0s3 c1t1d0s7 2.
Create these replicas on all hosts that use the shared diskset. Make sure that all the following conditions are satisfied for the creation-target disk: ¢ The disk is not managed by HDLM. ¢ The disk is not shared. The disk is not connected via Fibre Channel. The following shows an example of executing the command: ¢ # metadb -a -f c1t1d0s1 c1t1d0s3 c1t1d0s7 2. Create a shared diskset. The following shows an example of executing the commands: host1# metaset -s diskset1 -a -h host1 host2 3.
host1# mkdir /diskset1 host1# mount /dev/md/diskset1/dsk/d1 /diskset1 When using Sun Cluster: host1# mkdir -p /global/diskset1/d1 host2# mkdir -p /global/diskset1/d1 host1# mount -g /dev/md/diskset1/dsk/d1 /global/diskset1/d1 Setting Up VCS When VCS is installed on an HDLM host, you can register HDLM devices, or VxVM disk groups containing HDLM devices, as resource groups. For details on registering resource groups, refer to the VCS documentation.
d: The total number of all HDLM devices #1 If the number of paths differs depending on an HDLM device, use the maximum number for the calculation. #2 The I/O timeout value used for sending a notification to the sd or ssd driver depends on the timeout settings for switches or HBAs. Removing HDLM This section explains how to return the HDLM environment to the way it was before HDLM was installed, and describes each step of the process.
Preparations for HDLM Removal Performing Operations on HDLM-Managed Devices To perform operations on HDLM-managed disks: 1. If necessary, back up the contents of the HDLM-managed devices to tape or some other storage medium. 2. For the HDLM devices, collect all relevant information on the correspondence between the logical device file names and the internal LUNs of the LUs in the storage system.
6. Use the following command to unmount all mount points that are using HDLM: # umount mount-point 7. If HDLM devices are being used in Solaris 10 ZFS, execute the following command to export all ZFS storage pools that are using HDLM: # zpool export pool-name 8. Edit the /etc/vfstab file.
As the utility instructs, enter the name of the VxVM disk device group you want to unregister. If the HDLM device has been configured as a Quorum device, perform the following operations to remove the Quorum device settings. 1. Check the device ID of the Quorum device. The following shows an example of executing the command: # /usr/cluster/bin/scconf -pv | grep Quorum # /usr/cluster/bin/scdidadm -L 2.
Make sure that the node that is in the maintenance mode has the following status: ¢ ¢ The node is in the Offline status. Both the current Quorum votes present and Quorum votes possible are 0. The following shows an example of executing the command: host2# /usr/cluster/bin/scstat -q 9. If the cluster is a two-node configuration, enter the install mode. If the cluster consists of three or more nodes, this step is not necessary so go to step 10.
For details on how to remove the VCS settings, refer to the VCS documentation. Remove VxVM Settings If you are using both HDLM and VxVM on the same host, before you remove HDLM, perform the following to delete the HDLM devices from VxVM. Note If you are using Sun Cluster, before deleting HDLM devices from VxVM, remove the settings for the disk device group (see Remove Solaris Cluster Settings on page 3-186 of Preparations for HDLM Removal on page 3-185). For HDLM version 04-01 or earlier: 1.
When using a local metadevice 1. Execute the following command to unmount the file system The following shows an example of executing the command: host1# umount /mnt/filesystem1 2. Delete the local metadevice that is using the HDLM device. The following shows an example of executing the command to delete local metadevice d0. host1# metaclear d0 When using a shared diskset 1. Delete the shared metadevice. The following shows an example of executing the command.
The following shows an example of executing the command to delete SVM local volume d0. host1# metaclear d0 When using a shared diskset 1. Delete the shared volume. The following shows an example of executing the command. When not using Sun Cluster: host1# umount /diskset1 host1# metaclear -s diskset1 d1 When using Sun Cluster: host1# umount /global/diskset1/d1 host1# metaclear -s diskset1 d1 2. Delete the disk from the shared diskset. The following shows an example of executing the command.
¢ When removing HDLM in a Solaris 11 environment, be sure to execute /opt/DynamicLinkManager/bin/removehdlm. Do not use the pkg uninstall command to directly remove a Solaris package. To remove HDLM: 1. Log in to Solaris as a user with root permissions. 2. Execute the following command to remove HDLM: For Solaris 8, Solaris 9, or Solaris 10 # pkgrm DLManager For Solaris 11 # /opt/DynamicLinkManager/bin/removehdlm 3.
Then, execute the following command at the ok prompt on all nodes: boot -r Note While a node is starting, the message below may be output to the console or syslog: Could not stat: ../../devices/pseudo/dlmndrv@1/ dlmfdrv@0,0:c,raw path not loaded. This message is output because HDLM is removed and the HDLM devices are deleted.
A virtual disk backend is a location where virtual disk data is stored. A disk, disk slice, file, or volume (such as ZFS, SVM, and VxVM) can be used for the backend. In the control domain, register the install_disk file on a built-in disk as a virtual disk backend. The I/O domain recognizes this disk as a virtual disk. The device name of this virtual disk is /dev/[r]dsk/c0d0. In the I/O domain, register an HDLM device used in the storage system as a virtual disk backend.
To remove HDLM from a host in which LDoms have been set: 1. Log in to the control domain. 2. Execute the ldm stop command to stop the guest domain. For example, to stop a guest domain ldg2, execute the command as follows: # ldm stop ldg2 3. Execute the ldm unbind command to unbind the resource of the guest domain. For example, to unbind a resource of the guest domain ldg2, execute the command as follows: # ldm unbind ldg2 4.
For example, to remove virtual disks vdisk1 and vdisk2 from the guest domain ldg2, execute the command as follows: # ldm rm-vdisk vdisk1 ldg2 # ldm rm-vdisk vdisk2 ldg2 5. Execute the ldm rm-vdsdev command to stop exporting virtual disk backends from the service domain on the I/O domain to which HDLM will be installed.
9. On the control domain, execute the ldm add-vdisk command to assign the virtual disk backends to the guest domain. For example, to assign the virtual disk backends vol1 and vol2 exported from the service domain ldg1 to the guest domain ldg2, execute the command as follows: # ldm add-vdisk vdisk1 vol1@ldg1-vds0 ldg2 # ldm add-vdisk vdisk2 vol2@ldg1-vds0 ldg2 10. On the control domain, execute the ldm bind command to bind (connect) the resource of the guest domain ldg2.
Settings Needed After HDLM Removal VxVM Settings When using VxVM, execute the following command to confirm that the DEVICE column of the LUs previously managed by HDLM is displayed in the cWtXdYsZ format (for example, c6t0d10s2): # vxdisk list If the value is displayed in the cWtXdYsZ format, perform the following procedures: 1. Delete the /etc/vx/disk.info file. 2.
1. Remove the device ID information of devices that do not exist. The following shows an example of executing the command: # /usr/cluster/bin/scdidadm -C 2. Assign device IDs. The following shows an example of executing the command: # /usr/cluster/bin/scdidadm -r 3. Notify the server of the new device ID information. The following shows an example of executing the command: # /usr/cluster/bin/scdidadm -ui 4. Update the Sun Cluster global device namespace.
File System Settings For the file systems of the LUs that were managed by HDLM, specify settings that allow the Solaris sd or ssd driver to be mounted or specify settings that allow volume manager's volumes to be mounted after setup is performed again. After removing HDLM, the ZFS file system that was exported before the removal can be imported for use. Application Program Settings For the application programs that were using HDLM devices, specify the device file of the Solaris sd or ssd driver.
If you are unable to complete removal even though no programs other than HDLM are using HNTRLib2, contact your HDLM vendor or maintenance company if there is a maintenance contract for HDLM. Note If the log output directory set in HNTRLib2 was not the default directory, the log files will not be deleted during removal. In this case delete these files after removal.
¢ 3-202 Even if you attempt to remove HNTRLib2 when HNTRLib is installed, HNTRLib will not be removed. If no other programs are using HNTRLib, delete it manually.
4 HDLM Operation This chapter describes operating procedures for HDLM, including how to operate HDLM and the HDLM manager, and how to change the configuration of the operating environment.
Notes on Using HDLM This section explains the notes on using HDLM and using the environment in which HDLM is installed. Make sure that you read this section. Displaying Path Information • The path IDs displayed by using the dlnkmgr command's view operation differ depending on the order in which paths were detected when the host started. For this reason, you should always use the path name to specify a path.
The shaded portion indicates mapping from combinations of sd or ssd device types and instance numbers to HDLM device logical device file names. Storage System • Start up the storage system before you start up the hosts, so that Solaris can detect the storage system. • Do not change the vendor ID or product ID of a storage system. If you change either of them, HDLM will not be able to recognize the storage system.
If the host is started in single-user mode, the displayed path status indicates an error even when the actual path status is normal. Therefore, before you use HDLM devices in single-user mode, execute the procedure described in Initializing HDLM When the Host Is Started in Single-User Mode on page 4-4Operation in Single-User Mode on page 4-4. Upgrading Solaris In a local boot disk environment: When upgrading Solaris, first remove HDLM, and then upgrade Solaris.
3. Execute the following command to set the HDLM logical device file name in the HDLM filter driver: # /opt/DynamicLinkManager/bin/dlmsetdevname 4. Execute the following command to start the HDLM manager: # /etc/init.d/DLMManager start Tasks that Can Be Performed in Single-User Mode By executing the procedure described in Initializing HDLM When the Host Is Started in Single-User Mode on page 4-4, you can perform the following tasks in single-user mode while applications are inactive.
Maintenance Tasks on Devices Connected by Paths in the Boot Disk Environment Note the following points when performing maintenance on devices connected to the boot disk in a boot disk environment running Solaris: • Ensure that at least one path connected to the boot disk is online when you perform a maintenance task.
For details on the displayed items and their descriptions, see view (Displays Information) on page 6-34. Changing the Status of Paths This section explains how to change path statuses. Changing the Status of Paths to Online To change the status of paths to online: 1. Check the current status of the paths. To place paths online by specifying an HBA port, CHA port, or single path check the path name or PATH_ID used to manage the path.
KAPL01061-I 3 path(s) were successfully placed Online; 0 path(s) were not. Operation name = online # 3. Check to see if the statuses of all the applicable paths have changed. The following shows an example of executing the command: # /opt/DynamicLinkManager/bin/dlnkmgr view -path Changing the Status of Paths to Offline(C) To change the status of paths to Offline(C): 1. Check the current status of the paths.
# /opt/DynamicLinkManager/bin/dlnkmgr view -path Viewing LU Information This section explains how to display LU information by using the dlnkmgr command. To display LU information, execute the dlnkmgr command's view operation with the -lu parameter specified.
The following shows an example of executing the view operation: # /opt/DynamicLinkManager/bin/dlnkmgr view -drv PathID HDevName Device LDEV 000000 c6t5d0 sd74 9970/9980.15001.0670 000001 c6t5d1 sd75 9970/9980.15001.0671 000002 c6t5d2 sd76 9970/9980.15001.0672 000003 c6t5d3 sd77 9970/9980.15001.0673 000004 c6t5d4 sd78 9970/9980.15001.0674 000005 c6t5d5 sd79 9970/9980.15001.0675 000006 c6t5d0 sd80 9970/9980.15001.0670 000007 c6t5d1 sd81 9970/9980.15001.0671 000008 c6t5d2 sd82 9970/9980.15001.
# /opt/DynamicLinkManager/bin/dlnkmgr view -path Viewing and Setting Up the Operating Environment This section explains how to display and set up the HDLM operating environment. Viewing the Operating Environment To display the operating environment, execute the dlnkmgr command's view operation with the -sys and -sfunc parameters specified. The following shows an example of executing the command: # /opt/DynamicLinkManager/bin/dlnkmgr view -sys -sfunc HDLM Version : x.x.
• Load balancing • Path health checking • Automatic failback • Intermittent error monitoring • Dynamic I/O path control • Displaying the physical storage system information • Error log collection level • Trace level • Error log file size • The number of error log files • Trace file size • The number of trace files • Audit log data collection • Audit log facility • Number of times the same path can be used for load balancing • Number of times the same path can be used for exten
Permanent KAPL01001-I The HDLM command completed normally. Operation name = view, completion time = yyyy/mm/dd hh:mm:ss # For details on the displayed items and their descriptions, see view (Displays Information) on page 6-34. Updating the License This section explains how to update the license. To update the license, execute the dlnkmgr command's set operation with the -lic parameter specified. When the confirmation message is displayed, enter y to execute, or n to cancel the command.
Elog Level : 3 Elog File Size (KB) : 9900 Number Of Elog Files : 2 Trace Level : 0 Trace File Size(KB) : 1000 Number Of Trace Files : 4 Path Health Checking : on(30) Auto Failback : on(1) Intermittent Error Monitor : off Dynamic I/O Path Control : off(10) HDLM Manager Ver WakeupTime Alive x.x.x-xx yyyy/mm/dd hh:mm:ss HDLM Alert Driver Ver WakeupTime ElogMem Size Alive x.x.x-xx yyyy/mm/dd hh:mm:ss 4096 HDLM Driver Ver WakeupTime Alive x.x.
# Among the displayed items, HDLM Manager, HDLM Alert Driver, and HDLM Driver indicate the HDLM component information. Also, you can view information for each HDLM component. Execute the dlnkmgr command's view operation with the -sys and subsequent parameter specified.
# /etc/init.d/dlmmanager stop The stop script that was set up during HDLM installation runs, and stops the HDLM manager. Execute the following dlnkmgr command's view operation to confirm that the HDLM manager has stopped. # /opt/DynamicLinkManager/bin/dlnkmgr view -sys -msrv HDLM Manager Ver WakeupTime Dead KAPL01001-I The HDLM command completed normally. Operation name = view, completion time = yyyy/mm/dd hh:mm:ss # When the HDLM Manager column shows Dead, the HDLM manager is inactive.
• Changing the HBA mounting slot • Installing or removing the FCode on an HBA • Changing the LU default controller for a storage system • Adding, deleting, or changing a CHA port • Changing the topology of a storage system • Changing the Fibre Channel switches settings (including connection ports and zoning) • Changing a target ID • Changing the instance number of the HBA driver by changing the contents of the /etc/path_to_inst file • Switching the kernel mode • Inheriting logical device
• The -r parameter was specified when executing the dlmsetconf utility for creating the HDLM driver configuration definition file. • The LUN configuration was changed for the path to be used by HDLM to determine the logical device file name of the HDLM device. If the HDLM logical device file has been re-created by specifying the different file name, perform the following procedures: • Change the disk settings being used by the application programs.
Execute the following command to display the driver configuration information, and then refer to the instance field of the HDLM driver obtained in step 1 (the driver name of the HDLM driver is dlmfdrv). Identify the path used to determine the HDLM logical device file name of the LU from the FilterPath property. # /usr/sbin/prtconf -vP The following shows an example of executing the command.
3. Return to the HDLM boot disk environment, following the procedure described in Configuring a Boot Disk Environment on page 3-82. Dynamic Reconfiguration (DR) for Solaris In Solaris 8 or Solaris 9, you cannot use the Solaris dynamic reconfiguration (DR) functionality on HBAs that exist on a path managed by HDLM. To add, delete, or replace such HBAs, you need to restart the host after dynamic reconfiguration.
• Adding a path Add a path to an existing LU. The following operations are not supported in the dynamic reconfiguration of the HDLM device. To enable the following operations, restart the host with the reconfiguration option specified.
new LUs, execute the dlmsetconf utility, and then execute the dlmcfgmgr utility for managing the HDLM configuration. • Without restarting the system, when you add a new path to the LU that HDLM manages, the logical device file of the sd or ssd device corresponding to the path will not be created. When you want to use the newly added path in HDLM, execute the dlmsetconf utility and then restart the host, with the reconfiguration option specified.
• After you execute the dlmcfgmgr utility to add a new LU, when you execute the path status display command (the view -path operation), the HDevName of the added LU may be displayed as a hyphen (-). • If you do not specify a disk label, the dlmcfgmgr utility might take time to complete the processing when you execute the utility. Adding a New LU (When Not Using Solaris Cluster) To add a new LU: 1. Set up the storage system and switch so that Solaris recognizes the LU.
Figure 4-1 Example of executing the format command In the above execution example, when you specify an LU, the message Disk not labeled. Label it now? will be output if a disk label has not been specified. In this case, enter y and specify a disk label. 3. Execute the dlmsetconf utility for creating the HDLM driver configuration definition file to update the HDLM driver configuration definition file (/ kernel/drv/dlmfdrv.conf). # /opt/DynamicLinkManager/bin/dlmsetconf 4-24 4.
In environments supporting dynamic reconfiguration of HDLM devices, execute the dlmcfgmgr utility for managing the HDLM configuration. HDLM device will be reconfigured without restarting the host. # /opt/DynamicLinkManager/bin/dlmcfgmgr -a In environments not supporting dynamic reconfiguration of HDLM devices, restart the host, with the reconfiguration option specified. # reboot -- -r 7. Check the status of the path for accessing the added LU. The following shows an example of executing the command.
Have the HDLM drivers on all nodes in the cluster recognize the LU newly recognized by the node, and make the configuration the same for each node. To add a new LU: 1. Set up the storage system and switch so that Solaris recognizes the LU. 2. Check the LU newly recognized by the node and its disk label. You can check the LU and verify that the disk label is specified by using the format command. The following figure shows an example of the format command.
4. Make sure that the key used for disk reservation has been registered for the LU newly recognized by the node. Execute the following command on any node: # /usr/cluster/lib/sc/scsi -c inkeys -d the-name-of-the-logicaldevice-file-on-slice-2-of-the-LU-newly-recognized-by-the-node The following shows an example of executing the command: # /usr/cluster/lib/sc/scsi -c inkeys -d /dev/rdsk/ c3t50060E80033A9902d2s2 If the key is displayed, perform the steps beginning with step 5.
10. Use the following dlmsetconf utility to create the HDLM driver configuration definition file (/kernel/drv/dlmfdrv.conf) and the HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/config/ dlmfdrv.unconf). # /opt/DynamicLinkManager/bin/dlmsetconf The device mentioned in Devices Managed by HDLM on page 2-3 will be set as an HDLM-managed device.
Note The following message might be output to the console window or syslog while the node is running: Could not read symbolic link for: /dev/rdsk/ c3t50060E80033A9902d2s2 path not loaded This message is output when HDLM deletes a logical device file for an sd or ssd device managed by HDLM. Perform the procedure for updating the device ID information (steps 17 to 20) so that this message is not output next time the node starts. For details, refer to the Sun Cluster documentation. 14.
The following shows an example of executing the command. The shaded portion indicates the path accessing the LU recognized for the first time by the HDLM driver. 17. Delete the device ID information for devices that do not exist. Execute the following command: # /usr/cluster/bin/scdidadm -C 18. Allocate the device ID. Execute the following command: # /usr/cluster/bin/scdidadm -r 19. Notify the node of the most recent device ID information.
20. Update the Sun Cluster global device namespace. Execute the following command: # /usr/cluster/bin/scgdevs 21. When using, as a Quorum device, an HDLM device of an LU newly recognized for the first time by HDLM driver, set up the Quorum device on any node in the cluster. The following shows an example of executing the command: Check the device ID of the HDLM device.
¢ ¢ A new LU cannot be added via dynamic reconfiguration when you are using VxVM, because the host must be restarted in order to use (in VxVM) an LU that has been newly added to HDLM. Perform the procedure described in Adding a New LU By Restarting the Nodes (When Using Solaris Cluster) on page 4-25. For details on the precautions for the dlmcfgmgr utility, see The dlmcfgmgr Utility for Managing the HDLM Configuration on page 7-17. To add a new LU: 1.
page 3-185. For details on how to remove SDS or SVM disksets, see Remove SDS Settings on page 3-189 or Remove SVM Settings on page 3-190 of Preparations for HDLM Removal on page 3-185. 4. Make sure that the key used for disk reservation has been registered for the LU newly recognized by the node.
Installation of HDLM on page 3-45, and then execute the dlmsetconf utility. All sd or ssd devices connected to the host will be set as HDLM-managed devices. 7. If you do not want HDLM to manage one of the LUs newly recognized by the node, set up the LU so that it is excluded from HDLM management. For details, see Using the HDLM Device Unconfiguration Function When Performing a New Installation of HDLM on page 3-72.
11. Delete the device ID information for devices that do not exist. Execute the following command: # /usr/cluster/bin/scdidadm -C 12. Allocate the device ID. Execute the following command: # /usr/cluster/bin/scdidadm -r 13. Notify the node of the most recent device ID information. Execute the following command: # /usr/cluster/bin/scdidadm -ui 14. Update the Sun Cluster global device namespace.
# /usr/cluster/bin/scgdevs 15. When using, as a Quorum device, an HDLM device of an LU newly recognized for the first time by HDLM driver, set up the Quorum device on any node in the cluster. The following shows an example of executing the command: Check the device ID of the HDLM device.
• Changing an LU# • Changing the host LUN • Changing, adding, or deleting a CHA port • Changing a target ID • Adding or replacing a Fibre Channel switch • Replacing, deleting, or adding an HBA # To change an LU, we recommend that you add a new LU and then delete the existing one, rather than changing a LU in the storage system management program. After adding the new LU, move data as required from the LU you are deleting to the new LU.
For details, see Set Up Solaris Cluster on page 3-51 in Preparations for a New Installation of HDLM on page 3-45. 6. If you are using VCS, unregister the HDLM devices you are reconfiguring and VxVM disk groups containing those HDLM devices, from the resource groups. For details on how to remove the VCS settings, refer to the VCS documentation. 7. If you are using VxVM, unregister the HDLM devices associated with the LUs you are deleting.
Execute the format command to make sure that the reconfigured environment is recognized by Solaris. 13. Move to the /opt/DynamicLinkManager/config directory and then execute the following command. Specify a different value in n each time you reconfigure the environment. # cp -ip dlmfdrv.unconf dlmfdrv.unconfbakn 14. Execute the dlmsetconf utility to create the HDLM driver configuration definition file (/kernel/drv/dlmfdrv.
Execution example after a path is added by the addition of an HBA port: Check that a path for the same HBA port has been added: 36a41,44 > ### Device=ssd74 target=65664 lun=2 > ### Path=/pci@8,700000/SUNW,emlxs@1/fp@0,0/ ssd@w50060e80033a9938,2 > ### Device=ssd84 target=134012 lun=2 > ### Path=/pci@8,700000/SUNW,emlxs@1/fp@0,0/ ssd@w50060e80033a9939,2 52a61,64 > ### Device=ssd73 target=65664 lun=3 > ### Path=/pci@8,700000/SUNW,emlxs@1/fp@0,0/ ssd@w50060e80033a9938,3 > ### Device=ssd83 target=134012 lun=3 >
Use the following command to apply the settings in the HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/config/ dlmfdrv.unconf) to the HDLM driver configuration definition file (/ kernel/drv/dlmfdrv.conf): # /opt/DynamicLinkManager/bin/dlmsetconf -u 18. When you use VxVM, from VxVM, disable the devices that you do not need to access.
# /usr/cluster/bin/scconf -a -q globaldev=d4 For cluster configurations with two nodes, execute the following command to clear the install mode for the cluster: # /usr/cluster/bin/scconf -c -q reset 26. If you cleared the settings for a disk device group in step 5, perform registration for the disk device group. If you are using VxVM, see Linking VxVM and Solaris Cluster on page 3-174. If you are using SDS, see To Use a Shared Diskset on page 3-178 of Registering HDLM Devices on page 3-177.
For details on how to perform these settings, refer to the documentation for the HBA being used. You may need to restart the host, depending on the HBA being used, its drivers, and the path configuration of the deleted LU. 3. Execute the dlmsetconf utility for creating the HDLM driver configuration definition file to update the HDLM driver configuration definition file (/ kernel/drv/dlmfdrv.conf).
When an LU is deleted via dynamic reconfiguration, path IDs become inconsecutive, as shown in this example. If you reboot the host, path IDs might change because they are reconfigured during the reboot. Adding a Path to an Existing LU by Dynamic Reconfiguration Note ¢ Only data disks can be reconfigured dynamically. You cannot change the configuration of boot disks. This section describes how to dynamically add a path to an existing LU. 1.
pci@2/pci@0/pci@8/SUNW,qlc@0,1/fp@0,0/ssd@w50060e8010027a84,5 (ssd4127) online Apr 2 16:03:36 shiga devfsadmd[89]: [ID 937045 daemon.error] failed to lookup dev name for /pci@400/pci@2/pci@0/pci@8/ SUNW,qlc@0,1/fp@0,0/ssd@w50060e8010027a84,5 # In the above example, underlines indicate the log entries of the newly recognized path. 3. Use the dlmsetconf utility shown below to update the HDLM driver configuration definition file (/ kernel/drv/dlmfdrv.conf).
ssd@w50060e8010027a80,2 ssd4119 000041 Online pci@400/pci@2/pci@0/pci@8/SUNW,qlc@0/fp@0,0/ ssd@w50060e8010027a84,2 0001 c7t50060E8010027A80d3 ssd4112 000018 Online pci@400/pci@2/pci@0/pci@8/SUNW,qlc@0,1/fp@0,0/ ssd@w50060e8010027a80,3 ssd4116 000019 Online pci@400/pci@2/pci@0/pci@8/SUNW,qlc@0,1/fp@0,0/ ssd@w50060e8010027a84,3 ssd4113 000042 Online pci@400/pci@2/pci@0/pci@8/SUNW,qlc@0/fp@0,0/ ssd@w50060e8010027a80,3 ssd4117 000043 Online pci@400/pci@2/pci@0/pci@8/SUNW,qlc@0/fp@0,0/ ssd@w50060e8010027a84,3 0
pci@400/pci@2/pci@0/pci@8/SUNW,qlc@0/fp@0,0/ ssd@w50060e8010027a80,1 ssd12 000047 Online /devices/ pci@400/pci@2/pci@0/pci@8/SUNW,qlc@0/fp@0,0/ ssd@w50060e8010027a84,1 KAPL01001-I The HDLM command completed normally. Operation name = view, completion time = 2012/04/02 16:05:26 # In the above example, underlines indicate the status of the newly added path.
about the LUs that are to be deleted but still contains information about other LUs. 3. Execute the HDLM configuration management utility (dlmcfgmgr) to delete the path that was disconnected from the host from the HDLM driver. # /opt/DynamicLinkManager/bin/dlmcfgmgr -a The HDLM device is deleted. After it is deleted, the host is not restarted. 4. Execute the view -path operation of the HDLM command to review path statuses after the configuration change.
Changing an HDLM-managed Device to a Non-HDLM-Managed Device To change an HDLM-managed device to a non-HDLM-managed device: 1. Check the LDEV information of the LU to be removed from HDLM management. Use the following command to check the LDEV information: # /opt/DynamicLinkManager/bin/dlnkmgr view -drv PathID HDevName Device LDEV 000000 c7t0d0 sd46 9970/9980.15001.066A 000001 c7t0d0 sd47 9970/9980.15001.066A 000002 c7t0d0 sd48 9970/9980.15001.066A ... # KAPL01001-I The HDLM command completed normally.
Figure 4-5 Example of executing the format command In this example, the shaded portion indicates the path information of the LU to be managed by HDLM. In the above execution example, when you specify an LU, the message Disk not labeled. Label it now? will be output if a disk label has not been specified. In this case, enter y and specify a disk label. 2. Edit the HDLM driver unconfiguration definition file (/opt/ DynamicLinkManager/config/dlmfdrv.unconf).
Figure 4-6 The /opt/DynamicLinkManager/config/dlmfdrv.unconf file (before editing) Figure 4-7 The /opt/DynamicLinkManager/config/dlmfdrv.unconf file(after editing) The shaded portion indicate the text to be edited. Note Do not edit anything other than adding the hash mark (#) to the beginning of the LDEV property line. 3. Apply the contents that were edited in step 2 to the HDLM driver configuration definition file (/kernel/drv/dlmfdrv.conf).
Specifying Whether a Logical Unit Is To Be Managed by HDLM (When Using Solaris Cluster) Changing an HDLM-Managed Device to a Non-HDLM-Managed Device Note ¢ ¢ When you remove an HDLM-managed LU from management, perform the remove operations on all nodes in the cluster, and make the configuration the same on each node. When you remove an HDLM-managed LU from management, restart the host. To change an HDLM-managed device to a non-HDLM-managed device: 1. Log in to Solaris as a user with root permissions.
specifies that the LU indicated by that LDEV attribute is not to be managed by HDLM. The following figure shows examples of editing the HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/config/ dlmfdrv.unconf). Figure 4-8 The /opt/DynamicLinkManager/config/dlmfdrv.unconf file (before editing) Figure 4-9 The /opt/DynamicLinkManager/config/dlmfdrv.unconf file (after editing) The shaded portion indicates the text to be edited.
Figure 4-10 Result output by executing the dlmvxexclude utility Figure 4-11 Coding in the /etc/vx/vxvm.exclude file The shaded portion indicates the device not to be managed by HDLM and to be released from the VxVM disabled state. For details on the dlmvxexclude utility, see The dlmvxexclude Utility for Assisting Creation of the VxVM Configuration File on page 7-31. 7. Stop all nodes in the cluster. On the desired one node, execute the following command: # /usr/cluster/bin/scshutdown -g0 -y 8.
Note The following message might be output to the console window or syslog while the node is running: Unable to open '/dev/rdsk/c5t50060E80033A9902d2s2' This message is output when an HDLM-managed device has been changed to a non-HDLM-managed device. Perform the procedure for updating the device ID information (steps 12 to 15) so that this message is not output next time the node starts. For details, refer to the Sun Cluster documentation. 9.
You can check the LU by using the format command. The following figure shows an example of the format command. The shaded portion represents an LU not managed by HDLM. Figure 4-12 Example of executing the format command 12. Delete the device ID information for devices that do not exist. Execute the following command: # /usr/cluster/bin/scdidadm -C 13. Allocate the device ID.
16. When using a non-HDLM-managed device, as a Quorum device, set up the Quorum device in any node on the cluster. The following shows examples of executing the commands: Check the device ID of the device not to be managed by HDLM: # /usr/cluster/bin/scdidadm -L Set up the Quorum device: # /usr/cluster/bin/scconf -a -q globaldev=d4 Execute the following command to clear the install mode of a cluster when it consists of two nodes: # /usr/cluster/bin/scconf -c -q reset 17.
Figure 4-13 Example of executing the format command In the above execution example, when you specify an LU, the message Disk not labeled. Label it now? will be output if a disk label has not been specified. In this case, enter y and specify a disk label. 3. If setup is being performed for Quorum devices and disk device groups in the LU to be managed by HDLM, release the LU settings. When you unregister the SDS or SVM diskset, the diskset data will be deleted so first back up any necessary data.
Execute the following command on any node: # /usr/cluster/bin/scshutdown -g0 -y 6. Start all nodes in the cluster in non-cluster mode. At the ok prompt, execute the boot -x command. 7. Unregister the key for the LU to be managed by HDLM.
Figure 4-15 The /opt/DynamicLinkManager/config/dlmfdrv.unconf file (after editing) The shaded portion indicates the text to be edited. Note Do not make any modifications other than adding or deleting the hash mark (#) in front of the LDEV property line corresponding to the LU. 10. Apply the settings in the HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/config/dlmfdrv.unconf) to the HDLM driver configuration definition file (/kernel/drv/dlmfdrv.
message is not output next time the node starts. For details, refer to the Sun Cluster documentation. 14. Execute the dlnkmgr command's view operation on all nodes in the cluster, and display the status of each program. The following shows an example of executing the command: # /opt/DynamicLinkManager/bin/dlnkmgr view -sys HDLM Version : x.x.
17. Delete the device ID information for devices that do not exist. Execute the following command: # /usr/cluster/bin/scdidadm -C 18. Allocate the device ID. Execute the following command: # /usr/cluster/bin/scdidadm -r 19. Notify the node of the most recent device ID information. Execute the following command: # /usr/cluster/bin/scdidadm -ui 20. Update the Sun Cluster global device namespace. Execute the following command: # /usr/cluster/bin/scgdevs 21.
Set up the Quorum device: # /usr/cluster/bin/scconf -a -q globaldev=d4 Execute the following command to clear the install mode of a cluster when it consists of two nodes: # /usr/cluster/bin/scconf -c -q reset 22.
You can check the LU and verify that the disk label is specified by using the format command. Figure 4-16 Example of executing the format command on page 4-64 shows an example of executing the format command. The shaded portion represents the LU to be managed by HDLM. Figure 4-16 Example of executing the format command In the above execution example, when you specify an LU, the message Disk not labeled. Label it now? will be output if a disk label has not been specified.
If the key is displayed, place the LU under HDLM management by following the steps from step 5 of Changing a Non-HDLM-Managed Device to an HDLM-Managed Device (When the Node Must Be Restarted) on page 4-57 in Specifying Whether a Logical Unit Is To Be Managed by HDLM (When Using Solaris Cluster) on page 4-52.
6. Apply the settings in the HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/config/dlmfdrv.unconf) to the HDLM driver configuration definition file (/kernel/drv/dlmfdrv.conf): Execute the following command: # /opt/DynamicLinkManager/bin/dlmsetconf -u 7. Execute the dlmcfgmgr utility shown below to make the HDLM driver recognize the LU to be managed by HDLM: # /opt/DynamicLinkManager/bin/dlmcfgmgr -a 8. Check the status of the path accessing the LU to be managed by HDLM.
# /usr/cluster/bin/scdidadm -ui 12. Update the Sun Cluster global device namespace. Execute the following command: # /usr/cluster/bin/scgdevs 13. When using an HDLM-managed device, as a Quorum device, set the Quorum device on any node in the cluster.
Inheriting logical device names during storage system migration This subsection describes inheritance of logical device names during storage system migration. Note that if the volume manager is either SVM or VxVM, logical device names cannot be inherited during storage system migration.
switches, cables between hosts and the storage system devices, and the HBA driver. 6. Set up switches and other devices so that the logical units in the destination storage system are recognized by the OS. # cfgadm -c configure ap_id For the value to be specified for ap_id, see the listing of hardware information displayed by the cfgadm -al command. The following shows an example of executing the command. # cfgadm -al Ap_Id Occupant system:slot0 configured system:slot1 configured 7.
Then, execute the dlmsetconf utility with the -u parameter specified to update configuration definition files. # /opt/DynamicLinkManager/bin/dlmsetconf -u 11. Execute the dlmcfgmgr utility as shown below so that HDLM recognizes the added logical units. # /opt/DynamicLinkManager/bin/dlmcfgmgr -a 12. Check the status of the paths to the added logical units. # /opt/DynamicLinkManager/bin/dlnkmgr view -path 13. Configure the upper-level software for the added logical units and other items.
5 Troubleshooting This chapter describes how to properly use HDLM error information, and then how to resolve the problem, if an error occurs in HDLM. The resolutions for path errors, HDLM program errors, and other types of errors are all described separately. If you need technical support, see Getting help on page xvii.
Information collected by using the DLMgetras utility for collecting HDLM error information Immediately after an error occurs, execute the DLMgetras utility to collect HDLM error information, because restarting the machine may delete existing error information already collected by the DLMgetras utility. For details on the DLMgetras utility and the error information it collects, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3.
The path name indicates a physical path. When you modify the system configuration or replace a piece of hardware, you should check the path names to identify the physical paths that will be affected by the change. A path name consists of the following four elements, separated by periods: ¢ Host port number (hexadecimal) ¢ Bus number (hexadecimal) ¢ Target ID (hexadecimal) Host LU number (hexadecimal) The path name is also the same as the PathName displayed by the dlnkmgr command's view operation.
Figure 5-1 Error Location When the KAPL08022-E Message is Output The following figure shows the troubleshooting procedure when the KAPL08022-E message is output. Figure 5-2 Troubleshooting Procedure When a Path Error Occurs If an error occurs in a boot disk environment on a path connected to the boot disk, perform the procedure described in Actions to Take for a Path Error in a Boot Disk Environment on page 5-6.
The following shows the procedure for using the dlnkmgr command to handle a path error. Examining the Messages Examine the message that is output to syslog in the host by using applications or tools for monitoring messages. If the KAPL08022-E message is output, view the message to check the path in which the error has occurred. For details on the message, see Checking Error Information in Messages on page 5-2. Obtain Path Information Obtain path information.
If any path cannot be placed online due to an error, the KAPL01039-W message will appear. To ignore such paths and to continue processing, type y. To cancel the processing, type n. Recheck the statuses of the paths that cannot be placed online, and resolve the problem. Actions to Take for a Path Error in a Boot Disk Environment This section describes the action to take if an error occurs in a boot disk environment on a path connecting to the boot disk.
¢ The message output to the console by Solaris ¢ The message output to the console by the HBA driver If an error occurs on one or more of the paths to the boot disk HDLM outputs the KAPL08022-E message to the console, syslog, and error log after the HDLM manager has started.
Figure 5-3 Troubleshooting Procedure When a Program Error Occurs The following shows the procedure for handling a program error by using the dlnkmgr command. Examining the Messages Examine the messages that are output to the host syslog. If an error occurs in an HDLM program, a message other than KAPL08xxx is output to syslog. Check the output message. Messages with error level E (Error) or higher require corrective action.
# /opt/DynamicLinkManager/bin/dlnkmgr view -sys If the KAPL01012-E message appears as a result of executing the command The following shows the KAPL01012-E message: KAPL01012-E Could not connect the HDLM manager. Operation name = view If this message appears, start the HDLM manager. For details on how to start the HDLM manager, see Starting the HDLM Manager on page 4-15.
5-10 Troubleshooting Hitachi Dynamic Link Manager (for Solaris) User Guide
6 Command Reference This chapter describes the HDLM command (dlnkmgr) and its operations.
Overview of the HDLM Command dlnkmgr This section describes how to specify the HDLM command dlnkmgr and its subcommands (called operations in HDLM). Command format Enter the command using the following format: dlnkmgr operation [parameter [parameter-value]] dlnkmgr The command name. operation The type of operation entered after dlnkmgr. parameter A value required for an operation. parameter-value A value required for an operation parameter.
Operation Functions Dynamically deletes a path that is an HDLM-management target. For details, see delete (Deletes a Path Dynamically) on page 6-80. delete Note: ¢ ¢ Execute the command as a user with root permissions. To specify a value that contains a space in its parameter, enclose the entire value in double quotation marks (").
example, when you want to execute the command in a shell script or batch file. Example # /opt/DynamicLinkManager/bin/dlnkmgr clear -pdst -s KAPL01001-I The HDLM command completed normally. Operation name = clear, completion time = yyyy/mm/dd hh:mm:ss # To display the format of the clear operation -help Displays the format of the clear operation. Example # /opt/DynamicLinkManager/bin/dlnkmgr clear -help clear: Format dlnkmgr clear -pdst [-s] KAPL01001-I The HDLM command completed normally.
¢ set ¢ view ¢ monitor ¢ add ¢ delete If you do not specify any operations, the help operation displays all operations available for the dlnkmgr command. Examples Example 1 The following example shows how to display all the operations available for the dlnkmgr command. # /opt/DynamicLinkManager/bin/dlnkmgr help dlnkmgr: Format dlnkmgr { clear | help | offline | online | set | view | monitor | add | delete } KAPL01001-I The HDLM command completed normally.
Valid value AutoPATH_ID { 000000 - 999999 }(Decimal) HBA_ID { 00000 - 99999 }(Decimal) CHA_ID { 00000 - 99999 }(Decimal) help: Format dlnkmgr help { clear | offline | online | set | view | monitor | add | delete } KAPL01001-I The HDLM command completed normally.
|-pathid AutoPATH_ID |-device SCSI-device-name} [-s] To display the format of the offline operation /opt/DynamicLinkManager/bin/dlnkmgr offline -help Parameters To place paths offline -path Indicates that the target of the operation is a path managed by HDLM. This parameter is optional because offline is always used for paths, so it is assumed. Make sure that you specify the paths to be taken offline by using the hba, -hbaid, -cha, -chaid, -pathid, or -device parameter. -hba host-port-number.
# /opt/DynamicLinkManager/bin/dlnkmgr view -hba For details about how to execute the view operation to display the HBA port IDs, see To display HBA port information on page 6-70 in Parameters on page 6-36. Example The following example shows how to place offline all paths connected to the HBA port whose HBA-port-ID is 00001. # /opt/DynamicLinkManager/bin/dlnkmgr offline -hbaid 00001 KAPL01102-I All the paths which pass the specified HBA port will be changed to the Offline(C) status.
the Offline(C) status, enter y. Otherwise, enter n. [y/n]:y KAPL01061-I 2 path(s) were successfully placed Offline(C); 0 path(s) were not. Operation name = offline # -chaid CHA-port-ID Use this parameter to place offline, at one time, all paths that pass through a specific CHA port. The command will place offline all paths connected to the CHA port that has the specified CHA-port-ID. Paths that pass through a specific CHA port on a storage system recognized by the operating system will be offline.
-device SCSI-device-name Use this parameter to place offline all paths connected to the specified SCSI device. For SCSI-device-name, specify the value of Device displayed by executing the view operation as follows: # /opt/DynamicLinkManager/bin/dlnkmgr view -lu For details on how to execute the view operation and display the SCSI device name, see To display LU information on page 6-57 in Parameters on page 6-36. Only one case-sensitive parameter value can be specified.
# /opt/DynamicLinkManager/bin/dlnkmgr offline -help offline: Format dlnkmgr offline [-path] { -hba HBAPortNumber.
online (Places Paths Online) The dlnkmgr command's online operation places offline paths online. Specify the paths to be placed online by specifying an HBA port, CHA port, or single path. Instead of specifying a path, you can also specify a SCSI device name. Format To place paths online /opt/DynamicLinkManager/bin/dlnkmgr online [-path] [-hba host-port-number.
Specify the host port number and bus number of the target HBA port: the numbers are found in the PathName field displayed using the view operation. Enter a period between these two parameter values. For details on the view operation, see view (Displays Information) on page 6-34.
Specify the current AutoPATH_ID of the target path, which is displayed by using the view operation. For details on the view operation, see view (Displays Information) on page 6-34. Leading zeros can be omitted (000001 and 1 indicate the same AutoPATH_ID). Therefore, when the target AutoPATH_ID is 000000, enter 000000 or 0 for the parameter value. AutoPATH_IDs are re-assigned every time the host is restarted.
Specify the current AutoPATH_ID of the target path, which is displayed by using the view operation. For details on the view operation, see view (Displays Information) on page 6-34. Leading zeros can be omitted (000001 and 1 indicate the same AutoPATH_ID). Therefore, when the target AutoPATH_ID is 000000, enter 000000 or 0 for the parameter value. AutoPATH_IDs are re-assigned every time the host is restarted.
parameter specification. To change the status of a specific LU, use the lu and -pathid parameters to specify the path to the LU. To change the status of all the paths, specify only -dfha. A regular online operation is executed on the paths other than the P-VOL in an HAM environment. -lu -pathid AutoPATH_ID Specify management-target paths for each LU (P-VOL). The target LUs are the LUs that belong to a path ID that you specify in the -pathid AutoPATH_ID parameter.
HBA_ID { 00000 - 99999 }(Decimal) CHA_ID { 00000 - 99999 }(Decimal) KAPL01001-I The HDLM command completed normally. Operation name = online, completion time = yyyy/mm/dd hh:mm:ss # Reference Executing the view operation together with standard UNIX commands enables you to filter the path information listed for a specific HBA port or CHA port. For details on the view operation, see view (Displays Information) on page 6-34.
|-elfs error-log-file-size |-elfn number-of-error-log-files |-systflv trace-level |-systfs trace-file-size |-systfn number-of-trace-files |-pchk {on [-intvl checking-interval]|off} |-afb {on [-intvl checking-interval]|off} |-iem { on [-intvl error-monitoring-interval] [-iemnum number-of-times-error-is-to-occur] | off } |-lic |-audlog {on [-audlv audit-log-data-collection-level] [-category [[ss] [a] [ca]|all]]|off} |-audfac facility-value |-lbpathusetimes number-of-path-use-times |-expathusetimes number-of-p
Item name Default value Recommended value Number of error log files 2 2 Trace level 0: Do not output any trace. 0: Do not output any trace. Trace file size 1000 (kilobytes) 1000 (kilobytes) Number of trace files 4 4 Path health checking on on (Interval is 30 minutes.) (Interval is 30 minutes.) The recommended checking interval depends on the operating environment. Automatic failback on (1-minute checking interval) The recommended checking interval depends on the operating environment.
off: Disabled -lbtype {rr|exrr|lio|exlio|lbk|exlbk} Specify the algorithm to be used for load balancing. rr: The Round Robin algorithm exrr: The Extended Round Robin algorithm lio: The Least I/Os algorithm exlio: The Extended Least I/Os algorithm lbk: The Least Blocks algorithm exlbk: The Extended Least Blocks algorithm The type of algorithm specified by the -lbtype parameter remains stored in the system, even when you disable the load balancing function by specifying -lb off.
name = set, completion time = yyyy/mm/dd hh:mm:ss # -elfs error-log-file-size Specify a value from 100 to 2000000 (in kilobytes) for the size of the error log files. The target log files for which a file size can be set are HDLM manager logs (dlmmgrn.log (n indicates a file number from 1 to 16)). The specified value is applied for HDLM manager logs. By specifying both the log file size and the number of log files, you can collect up to 32,000,000 kilobytes (approximately 30GB) of error logs in total.
Specify the size of the trace file in kilobytes. Specify a value from 100 to 16000. By specifying both the trace file size and the number of trace files, you can collect up to 1024000 kilobytes of trace files in total. If a value smaller than the set value is specified, the execution confirmation message KAPL01097-W is displayed and the trace file is deleted for the time being. The trace files for which a file size can be set are hdlmtrn.log (n indicates a file number from 1 to 64).
When you change the checking interval, the new setting takes effect immediately. When the checking interval is shortened and the checking interval after the change (from the end of the previous path health check) has already elapsed, the path health check will start over. The path health check interval setting remains stored in the system even if you disable the function by changing the path health checking to off.
Specify the interval between path status checks. Specify a value from 1 to 1440 minutes. The default is 1. Specify an interval appropriate for your operating environment.
If you omit the intermittent error monitoring interval or the number of times that the error is to occur, each setting is specified as follows: ¢ When the intermittent error monitoring interval or the number of times that the error is to occur has not been specified before: The intermittent error monitoring interval is set to 30 minutes, and the number of times that the error is to occur is set to 3.
During intermittent error monitoring, if you change the number of times that the error is to occur, the error count and the elapsed time measured since monitoring starts are reset to 0, and monitoring starts by using the new settings. Outside the duration of intermittent error monitoring, if you change the number of times that the error is to occur, the new settings take effect after the next time automatic failback succeeds.
Setting AFB IEM Executable operation Set IEM to on. Change the IEM settings. Set IEM to off. off off on#2 off Result of operation • When a path is being monitored (during a period of conditional intermittent error monitoring), the value of the error count and the elapsed monitoring time are reset to 0, and then intermittent error monitoring will restart. • When a path is not being monitored, nothing changes.
#2 Because automatic failback is off, intermittent error monitoring is disabled. Example The following example shows how to monitor for intermittent errors. # /opt/DynamicLinkManager/bin/dlnkmgr set -iem on -intvl 20 iemnum 2 KAPL01049-I Would you like to execute the operation? Operation name = set [y/n]: y KAPL0#1001-I The HDLM command completed normally. Operation name = set, completion time = yyyy/mm/dd hh:mm:ss -lic Specify this option for when a license is updated.
A temporary license key cannot be installed by using the dlnkmgr command's set operation. Example 1 The following example shows how to update the license key when the license key file exists: # /opt/DynamicLinkManager/bin/dlnkmgr set -lic KAPL01049-I Would you like to execute the operation? Operation name = set [y/n]: y KAPL01071-I A permanent license was installed.
-category [[ss] [a] [ca]|all] Specifies the categories of audit log data to be collected. The table below lists and describes the values used for this setting. The default is all. If you specify the -category parameter without specifying a value, all is assumed. Table 6-8 Values Indicating Audit Log Data Categories Value Explanation ss Audit log events of the StartStop category are collected. a Audit log events of the Authentication category are collected.
Specifies the number of times the same path can be used for sequential I/O operations when the extended Round Robin (exrr), Least I/Os (exlio), or Least Blocks (exlbk) algorithm is used for load balancing. You an specify a decimal (base 10) value from 0 to 999999. The default is 100. If you specify 0, the same path is used as long as the sequential I/O operations continue.
Table 6-10 Display items for which the display results of the view operation differ depending on the -pstv parameter specification Operation view -path Display Item DskName iLU ChaPort (CP) view -lu Product SerialNumber (S/N) iLU ChaPort view -drv LDEV The set -pstv setting does not affect the display results of the view cha operation. Information about the physical storage system is always displayed.
| -iem off | -lic | -audlog on [ -audlv AudlogLevel ] [ -category Category-Value ] | -audlog off | -audfac { Facility-Name | Facility-Number } | -lbpathusetimes Number-Of-PathUseTimes | -expathusetimes Number-Of-ExPathUseTimes | -dpc { on | off } [-pathid AutoPATH_ID { lu | -storage } ] | -dpcintvl Dpc-Interval | -pstv { on | off } } [-s] Valid value ElogLevel { 0 | 1 | 2 | 3 | 4 } (Default Value 3) ElogFileSize { 100 - 2000000 }(KB) (Default Value 9900) Number-Of-ElogFiles { 2 - 16 }(Files) (Default Value
yyyy/mm/dd # hh:mm:ss view (Displays Information) The dlnkmgr command's view operation displays HDLM program information, path information, LU information, HBA port information, CHA port information, and corresponding information about an HDLM device, sd or ssd device, and LDEV.
[-pstv|-vstv] [-stname] [-srt {lu|cp}] [-t] To display LU information To display LU information /opt/DynamicLinkManager/bin/dlnkmgr view -lu [-pstv|-vstv] [-hdev host-device-name | -pathid AutoPATH_ID] [-t] To display LU information (by adding items to be displayed) /opt/DynamicLinkManager/bin/dlnkmgr view -lu -item [ [slpr] [pn] [cp] [clpr] [type] [ic] [ie] [dnu] [iep] [pd] [dpc] [phys] [virt] [vid] [ha] [hastat] |all ] [-pstv|-vstv] [-hdev host-device-name|-pathid AutoPATH_ID] [-t] To display a summary
To display the format of the view operation /opt/DynamicLinkManager/bin/dlnkmgr view -help Parameters This section describes the parameters for the view operation, in the following order: To display program information on page 6-36 To display path information on page 6-43 To display LU information on page 6-57 To display HBA port information on page 6-70 To display CHA port information on page 6-71 To display corresponding information about an HDLM device, sd or ssd device, and LDEV on page 6-72 To display
Parameter and program information to be displayed HDLM function settings Item Load Balance Description Settings for load balancing • Setting status: on: Enabled off: Disabled • Algorithm: When the setting status of load balancing is on, one of the following types of algorithms is used for load balancing is displayed in the parentheses following on.
Parameter and program information to be displayed Item Description • 4: Outputs all information. Trace File Size(KB) Trace file size in kilobytes Number Of Trace Files Number of trace files. Path Health Checking • Settings for path health checking: on: Enabled off: Disabled • Checking interval: When the setting of the path health checking is on, the checking interval of path health checking is displayed within the parentheses following on. The time is in minutes.
Parameter and program information to be displayed Item Description • Checking interval The parentheses following the setting status shows the checking interval for reviewing information about the switching of controllers performed by the storage system. "Minute" is used as the unit. If different settings have been specified for each storage system or LU, an asterisk (*) is added after the parentheses in which the checking interval is displayed.
Parameter and program information to be displayed Item Description For example, when there are 100 days left until the license period (2006/08/21) expires, the following appears: 2006/08/21(100days after) -audlog Audit Log Settings for audit log data collection: • Information about audit log data collection settings Whether collection is enabled: on: Enabled off: Disabled • Audit log data collection level: When audit log data collection is on, the collection level that has been set is displayed wit
Parameter and program information to be displayed Item extended load balancing -pstv The display-ofthe-physicalstoragesysteminformation setting Description If you used Global Link Manager to set the number of times the same path can be used in units of LUs, an asterisk (*) is added after the value. Physical Storage View The value of the display-of-the-physical-storagesystem-information setting is displayed.
= view, completion time = yyyy/mm/dd hh:mm:ss # Example 3 The following example shows how to display information about the HDLM alert driver: # /opt/DynamicLinkManager/bin/dlnkmgr view -sys -adrv HDLM Alert Driver Ver WakeupTime ElogMem Size Alive x.x.x-xx yyyy/mm/dd hh:mm:ss 4096 KAPL01001-I The HDLM command completed normally.
# Example 8 The following example shows how to display the number of times the same path can be used for extended load balancing: # /opt/DynamicLinkManager/bin/dlnkmgr view -sys -expathusetimes Times Same ExPath Was Used : 100 KAPL01001-I The HDLM command completed normally.
displayed according to the value specified in the -pstv parameter of set operations. For information about the items that are displayed differently depending on the specified -pstv and -vstv parameters, see Table 6-10 Display items for which the display results of the view operation differ depending on the -pstv parameter specification on page 6-32 in Parameters on page 6-18. -hdev host-device-name Displays information only for the paths accessing the specified host device.
Omits the title for each information item. Example The following example shows how to display information about the paths that access a host device. To display path information (by selecting a display item) -path -item When you specify the -path parameter together with the -item parameter, the command only displays the items specified by the value of the -item parameter. If you specify no value for the -item parameter, only the PathID and the Status fields are displayed.
Selectable items Sub-parameters DNum dnu HDevName hd IEP iep HBAPortWWN hbaportwwn Physical-LDEV phys Virtual-LDEV virt Physical-DskName vid Physical-iLU vid Physical-ChaPort vid Org-DskName ha Org-iLU ha # Because both PathID and Status are always displayed, you do not have to specify any parameters. In the sub-parameters following -path -item, you can list the paths (hdev) and sort the list (-srt).
-stname Use this parameter to display the model ID of the storage system in the product ID element of the DskName field. When this parameter is omitted, the command displays the product ID or emulation type of the storage system instead. For details about the information to be displayed for product IDs, see Table 6-15 Product ID Displayed by the view -path Operation on page 6-56. When you use this parameter, DskName is displayed by default. It is not necessary to specify dn for the -item parameter.
To display a summary of path information -path -c When you specify the -path parameter together with the -c parameter, only the PathID, DskName, iLU, CP, Status, and Type fields are displayed from the path information managed by HDLM. The displayed contents are shortened so that each instance of path information is displayed on a single line. For details on what is displayed for in each item, see Table 6-13 Displayed Path Information on page 6-50.
displayed according to the value specified in the -pstv parameter of set operations. For information about the items that are displayed differently depending on the specified -pstv and -vstv parameters, see Table 6-10 Display items for which the display results of the view operation differ depending on the -pstv parameter specification on page 6-32 in Parameters on page 6-18. -stname Use this parameter to display the model ID of the storage system in the product ID element of the DskName field.
Displayed path information Table 6-13 Displayed Path Information on page 6-50 describes the displayed path information. The following explains the table headings: • No summary displayed: The user specifies only the -path parameter or the -path -item parameter. • Summary displayed: The user specifies the -path -c parameter. Table 6-13 Displayed Path Information Displayed Item No Summary summary displayed displayed Description Paths Total number of displayed paths, indicated by a decimal number.
Displayed Item No Summary summary displayed displayed Description IO-Count Total I/O count for all the displayed paths, indicated by a decimal number. The maximum value that can be displayed is 232 - 1 (4294967295). If the total I/O count reaches the maximum value, it will re-start from 0. IO-Errors Total I/O error count for all the displayed paths, indicated by a decimal number. The maximum value that can be displayed is 232 - 1 (4294967295).
Displayed Item No Summary summary displayed displayed Description This number combined with the storage system name (shown in DskName) identifies the LU that is accessed by a path. ChaPort#1 CP • For the Lightning 9900V Series, Hitachi NSC55, Hitachi Universal Storage Platform 100, Hitachi Universal Storage Platform 600, Hitachi Universal Storage Platform 1100, or HUS VM indicated by a hexadecimal number.
Displayed Item No Summary summary displayed displayed Description • Online(D): The paths to the primary volume (P-VOL) in an HAM environment can be recovered automatically. Paths that are Offline(E) or Online(E) require corrective action. Take appropriate action by referring to What To Do for a Path Error on page 5-3.
Displayed Item No Summary summary displayed displayed Description One of the following values is displayed for each path: • Indicates that intermittent error monitoring is disabled or the monitoring time for an intermittent error is out of range (the path status is Online(E) or Offline(E)). • A value of at least 0 Indicates the number of errors that occurred during intermittent error monitoring (the path status is Online(E), Offline(E), or Online).
Displayed Item No Summary summary displayed displayed Description • For HUS VM indicated by a hexadecimal number. The first two characters of iLU are the CU number, and the last two characters are the internal LU number within the CU. • For Hitachi Virtual Storage Platform, indicated by a hexadecimal number. The first two characters of iLU are the number of the logical DKC (Disk Controller), the middle two numbers are the CU number, and the last two characters are the internal LU number within the CU.
#1 The path information is displayed only when a value is specified for the path -item parameter. #2 This storage system applies when the dynamic I/O path control function is disabled. Table 6-14 Elements of a Path Name Element Host port number Corresponding information Solaris representation Host port number A unique number in a host: HDLM assigns this unique number to each HBA port. When two HBAs that have two ports are installed on the host, a value from 0000 to 0003 is assigned to each port.
Product ID Model names of storage systems • Without the -stname parameter With the -stname parameter (Displays the following for the model ID) Hitachi Universal Storage Platform 1100 SVS SVS • Hitachi Universal Storage Platform V • Hitachi Universal Storage Platform VM • Hitachi Virtual Storage Platform USP_V VSP VSP G1000 VSP_G1000 HUS VM HUS_VM • P9500 P9500 XP7 XP7 XP128 XP128 XP1024 XP1024 XP10000 XP10000 XP12000 XP12000 XP20000 XP20000 XP24000 XP24000 # When a summar
displayed. The sorting key is iLU and its configuration information is displayed for each LU. For details on the content of each displayed item, see Table 6-18 Displayed LU Information on page 6-64. The following shows the sub-parameters: -pstv|-vstv If the -pstv parameter is specified, information about the physical storage system is displayed. If the -vstv parameter is specified, information about the storage system recognized by the operating system is displayed.
0671 c6t5d1 sd75 000001 Online sd81 000007 Offline(E) 0672 c6t5d2 sd76 000002 Online sd82 000008 Offline(C) KAPL01001-I The HDLM command completed normally. Operation name = view, completion time = yyyy/mm/dd hh:mm:ss # To display LU information (by adding items to be displayed) -lu -item The items specified with the -item parameter are added to the items to be displayed by the -lu parameter and displayed.
Item that can be added Sub-parameter Physical-ChaPort vid Org-Product ha Org-SerialNumber ha Org-iLU ha HaStat hastat All items are displayed all For details on the contents of each displayed item, see Table 6-18 Displayed LU Information on page 6-64. The following shows the sub-parameters: -pstv|-vstv If the -pstv parameter is specified, information about the physical storage system is displayed.
Example 2 When using Hitachi AMS2000 series, Hitachi SMS series, or HUS100 series and displaying LU information with DPC added to the display items: # /opt/DynamicLinkManager/bin/dlnkmgr view -lu -item dpc Product : HUS100 SerialNumber : 9203008 LUs : 3 Dynamic I/O Path Control : on* iLU HDevName DPC PathID Status 000006 c6t5d0 on 000000 Online 000003 Online 000007 c6t5d1 off 000001 Online 000004 Online 000008 c6t5d2 on 000002 Online 000005 Online # To display a summary of LU information -lu -c Command Re
When the -c parameter is specified with the -lu parameter, a summary of LU configuration information is displayed on one line. The total number of paths recognized by HDLM and the number of online paths are displayed for each LU. You cannot specify the -c parameter together with the -hdev or -pathid parameter. For details on the contents of each display item, see Table 6-18 Displayed LU Information on page 6-64.
The following table lists the correspondence between the display items that can be added by using the -item parameter and the parameter values that can be specified after the -item parameter. Table 6-17 Items That Can Be Added by Using the -lu -c -item Parameter and Sub-parameters Item that can be added SLPR Sub-parameter slpr The following show the sub-parameter: -pstv|-vstv If the -pstv parameter is specified, information about the physical storage system is displayed.
Table 6-18 Displayed LU Information Displayed item No summary displayed Model ID of the storage system Product Serial number Description Summary displayed S/N Serial number of the storage system Total number of LUs managed by HDLM among the LUs in the storage system LUs Dynamic I/O Path Control -- The setting information about the dynamic I/O path control function is displayed for each storage system. on: The dynamic I/O path control function is enabled.
Displayed item No summary displayed Description Summary displayed the storage system. You can identify an actual LU by referencing iLU from the storage system management program. • For Universal Storage Platform V/VM series, Hitachi Virtual Storage Platform, or VSP G1000 indicated by a hexadecimal number. The first two characters of iLU are the number of the logical DKC (Disk Controller), the middle two numbers are the CU number, and the last two characters are the internal LU number within the CU.
Displayed item No summary displayed Description Summary displayed • Host port number (hexadecimal number) • Bus number (hexadecimal number) • Target ID (hexadecimal number) • Host LU number (hexadecimal number) For details on each element of the path name and its representation in Solaris, see Table 6-14 Elements of a Path Name on page 6-56 ChaPort#1 -- Port number of the CHA, which identifies the CHA port that is mounted on the storage system.
Displayed item No summary displayed Description Summary displayed series, Hitachi AMS2000 series#3, Hitachi SMS series#3, HUS100 series#3, or HUS VM all paths are owner paths. IO-Count#1 -- Total I/O count for a path. The maximum value that can be displayed is 232 - 1 (4294967295). If the total I/O count reaches the maximum value, it is reset, and the count is re-started from 0. To reset the IO-Count value to 0, execute the dlnkmgr command's clear operation.
Displayed item No summary displayed Description Summary displayed You can identify the physical volume from this information. If the volume is not virtualized, a hyphen (-) is displayed. VirtualLDEV -- Displays the model ID, serial number, and iLU number of a virtual volume, separated by periods. You can identify the virtual volume from this information. If the volume is not virtualized, a hyphen (-) is displayed.
Displayed item No summary displayed Description Summary displayed paths and take resolve the problem for any paths that have an error status. Org-Product -- For HAM environments, the model ID of the storage system on the secondary volume (S-VOL) is displayed. If an HAM environment is not used, a hyphen (-) is displayed. OrgSerialNumbe r -- For HAM environments, the serial number of the storage system on the secondary volume (S-VOL) is displayed.
This storage system applies when the dynamic I/O path control function is disabled. To display HBA port information -hba Use this parameter to display information about the paths that pass through each HBA port. HDLM displays the unique HBA port ID assigned to each HBA port. This HBA port ID is used in monitor operations, and can also be used in Offline and Online operations. The following table lists and describes the displayed items.
name = view, completion time = yyyy/mm/dd hh:mm:ss # The following shows the subsequent sub-parameters: -srt pb Sorts the HBA port information by the host port numbers (Port) and bus numbers (Bus). -c Does not display the SCSI device configuration information (Physical Device). -portwwn Displays port WWN information for HBAs connected to the storage system using a 16-digit hexadecimal number. -t Omits the title for each information item.
Item OnlinePaths Description Number of available paths in the displayed paths, indicated by a decimal number. When the value of Paths equals the value of OnlinePaths, all paths are online. If the value of OnlinePaths is less than that of Paths, some paths might have an error status, in which case you should check the status of individual paths and take appropriate action for any paths that have an error status.
-t Omits the title for each information item. Table 6-21 To Display Corresponding Information About an HDLM Device, sd or ssd Device, and LDEV Item PathID Description AutoPATH_ID, indicated by a decimal number. AutoPATH_ID is assigned when the host is restarted. When a new LU was added and the host has not been restarted, AutoPATH_ID is re-assigned to each path of the LU when you execute the dlmcfgmgr utility for managing the HDLM configuration. HDevName Host device name.
# /opt/DynamicLinkManager/bin/dlnkmgr view -help view: Format dlnkmgr view -sys [ -sfunc | -msrv | -adrv | -pdrv | -lic | -audlog | -lbpathusetimes | -expathusetimes | -pstv ] [-t] dlnkmgr view -path [-pstv | -vstv] [ -hdev HostDeviceName ] [-stname] [-iem] [-srt {pn | lu | cp}] [-hbaportwwn] [-t] dlnkmgr view -path -item [pn] [dn] [lu] [cp] [type] [ic] [ie] [dnu] [hd] [iep] [hbaportwwn] [phys] [virt] [vid] [ha] [-pstv | -vstv] [-hdev HostDeviceName] [stname] [-srt {pn | lu | cp}] [-t] dlnkmgr view -path -c
If you perform dynamic reconfiguration of the HDLM device during the monitor operation, the KAPL01078-W message is displayed and the monitor operation stops.
Table 6-22 Recommended values for the -intvl parameter when there are 500 or more paths Number of paths (approximate) Recommended value specified for the -intvl parameter (seconds) 500 3 or more 1000 5 or more 2000 10 or more 4000 30 or more To display I/O information for each HBA port -hbaid HBA-port-ID Specifies the HBA port ID of the HBA port for which I/O information is to be displayed. The subsequent parameters specify the display interval and count.
Item Description (4294967295). If the total I/O count reaches the maximum value, it is reset, and the count is re-started from 0. IO-Errors Total I/O error count for the displayed paths during a specified interval, indicated by a whole number (indicated by a decimal number). The maximum value that can be displayed is 232 - 1 (4294967295). If the total I/O error count reaches the maximum value, it is reset, and the count is re-started from 0.
Table 6-24 Items displayed as I/O information for each CHA port Item Description ChaID CHA port ID Product Model ID of the storage system SerialNumber Serial number of the storage system ChaPort Port number of the CHA, which identifies the CHA port that is mounted on the storage system. You can identify an actual CHA port by referencing this number from the storage system management program.
dlnkmgr monitor -hbaid HBA_ID [-intvl Interval-Time] [-count Number-OfTimes] [-t] dlnkmgr monitor { -cha -pathid AutoPATH_ID | -chaid CHA_ID } [-intvl Interval-Time] [-count Number-OfTimes] [-t] Valid value HBA_ID { 000000 - 99999 }(Decimal) CHA_ID { 00000 - 99999 }(Decimal) Interval-Time { 1 - 60 }(Seconds) (Default Value 3) Number-Of-Times { 1 - 999 }(Times) (Default Value 10) KAPL01001-I The HDLM command completed normally.
-s Executes the command without displaying the message asking for confirmation of command execution from the user. Specify this parameter if you want to skip the response to the confirmation message: for example, when you want to execute the command in a shell script or batch file. Example # /opt/DynamicLinkManager/bin/dlnkmgr add -path -s KAPL01162-I A path was added. (path ID = 00010, storage = HITACHI.HUS100.9100163, iLU = 0030) : KAPL01159-I Paths were added.
Format To Delete a Path Dynamically /opt/DynamicLinkManager/bin/dlnkmgr delete -path [-s] To Display the Format of the delete Operation /opt/DynamicLinkManager/bin/dlnkmgr delete -help Parameters To Delete a Path Dynamically -path Indicates that the target of the operation is a path managed by HDLM. Example # /opt/DynamicLinkManager/bin/dlnkmgr delete -path KAPL01161-I This operation will change the path configuration. Do you want to continue? [y/n]:y KAPL01165-I A path was deleted.
• Set up the storage system (remove the mapping between the internal LU and the storage port LUN). • Terminate the connection between the storage port and the host HBA port. • Release the Fibre Channel switch settings. • Release the settings for the host HBA and for the sd or ssd driver. For details on how to perform these settings, refer to the documentation for the HBA being used.
7 Utility Reference This chapter explains the utilities used by HDLM.
Overview of the Utilities HDLM provides the following utilities: • The DLMgetras utility for collecting HDLM error information When an error occurs, this utility collects the files that contain information to be submitted to your HDLM vendor or maintenance company. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3.
This utility outputs the file that has the same format as the VxVM management target exclusion file that matches the definition in the HDLM driver configuration definition file (/kernel/drv/dlmfdrv.conf) to the specified directory. For details on the dlmvxexclude utility, see The dlmvxexclude Utility for Assisting Creation of the VxVM Configuration File on page 7-31.
When the host is restarted, a part of the information that was collected by the DLMgetras utility will be cleared. If an error occurs, immediately execute this utility.
¢ ¢ ¢ Do not specify a directory that contains the directory to which the collected information is output. If you specify this directory, the DLMgetras utility will run indefinitely. Lines beginning with a hash mark (#) are handled as comment lines. If a hash mark (#) is encountered anywhere other than at the beginning of a line, it is assumed to be part of the path name. ¢ Only one file or directory can be specified per line. ¢ The root directory (/) cannot be specified.
List of Collected Error Information The following illustrates the error information collected by executing the DLMgetras utility, which is explained separately in each output file. • Name-of-the-directory-containing-a-system-crash-dump-file/name-ofthe-system-crash-dump-file This is a symbolic link file to a system crash dump file. Set all files in the directory specified by the DUMPADM_SAVDIR parameter that is in the /etc/dumpadm.conf file. • hbsa.tar.
Output directory#1 Files dlmcfgmgr1.log dlmcfgmgr2.log Explanation Execution log file of the dlmcfgmgr utility which collects the details of error occurred during the dlmcfgmgr utility for managing the HDLM configuration execution hdlmtr[1-64].log Trace file installhdlm.log Execution log file of the installhdlm utility for installing HDLM dlmstart[1-2].log Execution log file of the dlmstart utility for Configuring HDLM Devices var/opt/ DynamicLinkManager/log/ mmap hdlmtr.
Output directory#1 Files Explanation opt/hitachi/HNTRLib/mmap hntrmmap.mm Memory mapped file created when HDLM 04-00 or earlier is used opt/hitachi/HNTRLib2/ mmap hntr2mmap.mm Memory mapped file opt/hitachi/HNTRLib/ spool hntr[1-16].
Output directory#1 Files bfa.conf, emlxs.conf, Explanation HBA driver setup file (file names depend on the installed HBA driver) fca-pci.conf, fcaw.conf, fcp.conf, fjpfca.conf, fjfcif.conf, fp.conf, jnic.conf, jnic146x.conf, jfca.conf, lpfc.conf, lpfs.conf, qla2200.conf, qla2300.conf, or qlc.conf etc md.conf Upper limit information on metadevice, metaset scsi_vhci.conf Sun StorageTek Traffic Manager Software setup and configuration definition file sd.
Output directory#1 Files Explanation mnttab File system mount information syslog.conf syslog output directory definition rsyslog.
Output directory#1 Files Explanation dlmgetomtrace.dmp In-memory trace information dlmls-la.txt List of directories and files related to the HDLM installation file information (output results from the ls -alR command) dlmmgr-cha.txt Output results of dlnkmgr view -cha dlmmgr-hba.txt Output results of dlnkmgr view -hba -portwwn dlmmgr-lu.txt HDLM LU information dlmmgr-lu-all.txt HDLM LU information (including the number of times an intermittent error occurred) dlmmgr-path.
Output directory#1 7-12 Files Explanation pkginfo-l.txt List of package information pkg-info.txt List of package information (for Solaris 11) pkg-history-l.txt List of package installation history (for Solaris 11) pkg-contentsDLManager.txt List of installed files (for Solaris 11) prtconf-pv.txt Hardware configuration information of a host (output results of the prtconf -pv command) prtconf-vP.txt List of driver information prtconfMemory.txt Physical memory size prtvtoc.
Output directory#1 VxVM #6 Files Explanation what.txt Output of what command zfs-list.txt#6 ZFS file system information (output results of the zfs list command) zpool-list.txt#6 Configuration information for ZFS storage pools (output results of the zpool list command) zpool-status.txt#6 Status information for ZFS storage pools (output results of the zpool status command) (Not applicable) VxVM information collection directory VXVM_pkginfo.txt VxVM package information VXVM_vxdmp.
Output directory#1 Files Explanation system-hdlmmanager:default.log. [0-9] system-hdlmmanager:default.log system-hdlminitpath:default.log. [0-9] Log of the HDLM Manager service (Solaris 10 or Solaris 11) system-hdlmLog of the hdlmonlinepath:default.log onlinepath service (Solaris 10 or Solaris 11) system-hdlmonlinepath:default.log .[0-9] var/vxvm#6 vxconfigd.log.txt Log of error messages The output destination is either /var/vxvm/ vxconfigd.log or /var/adm/syslog/ syslog.
Output directory#1 var/VRTSvcs/log#6 SDS#6 etc/lvm#6 SVM#6 etc/opt/SUNWmd#6 Files Explanation VCS_hasys-list.txt System information in the VCS cluster VCS_hagrp-display.txt Service group information VCS_hares-display.txt Resource information VCS_hasys-display.txt VCS settings engine_[A, ...].log VCS engine log (Not applicable) SDS information collection directory SDS_pkginfo.txt SDS package information SDS_metadb.txt Status database information SDS_metastat.
Output directory#1 SC3.0#6 Files (Not applicable) Explanation Sun Cluster information collection directory SC3.0_pkginfo.txt Package information SC3.0_scstat-q.txt Cluster status snapshot SC3.0_scconf-p.txt Cluster configuration information SC3.0_scdidadm-L.txt Mapping information between HDLM drivers and device IDs SC3.0_dev_did_dsk.txt List of directories and files in the /dev/did/dsk directory SC3.0_dev_did_rdsk.
The dlmcfgmgr Utility for Managing the HDLM Configuration This utility makes the HDLM driver recognize the contents of the HDLM driver configuration definition file (/kernel/drv/dlmfdrv.conf) dynamically without restarting the host. The following table lists the operations to be performed and not to be performed.
Parameters -a Applies the contents of the HDLM driver configuration definition file (/ kernel/drv/dlmfdrv.conf) to HDLM driver. The added LU becomes valid after the command with this option is executed. The HDLM devices for the deleted LUs become unavailable after the command with this option is executed. -h Displays help about the usage of the command.
¢ ¢ ¢ ¢ ¢ ¢ ¢ If the HDLM driver configuration definition file (/kernel/drv/ dlmfdrv.conf) contains the definition of a newly added LU, the dlmcfgmgr utility makes the HDLM driver recognize the new LU so that the LU is to be managed by HDLM. If the definition of an existing LU no longer exists in the HDLM driver configuration file (dlmfdrv.conf file), the dlmcfgmgr utility deletes the existing LU from the HDLM driver and excludes it from HDLM management.
¢ ¢ ¢ If no change is made to the HDLM-managed devices when the dlmcfgmgr utility is executed, the utility outputs the message and terminates. You cannot concurrently execute more than one dlmcfgmgr utility on the same host. If you attempt such execution, the utility outputs the message and terminates. When any of the operations below are performed in a Solaris 10 or Solaris 11 environment, a message in the format below is output from the devfsadmd daemon to syslog or the console. Ignore that message.
Format /opt/DynamicLinkManager/bin/dlmlisthdev input-file-name Parameters input-file-name Specifies, as an input file, the name of the output file that stores the output result of the dlnkmgr command's view -path operation executed before HDLM 04-00 or 04-01 was removed. You cannot specify multi-byte characters. Examples Example Referring to the output results of the dlnkmgr command's view -path operation, which are saved in the /var/tmp/old_view_path_list.
in HDLM 04-00 and those in the newer version of HDLM, and displays the result in the standard output. The result of the command execution is as follows: The following table lists and describes the items to be displayed. Table 7-3 Information Displayed by the dlmlisthdev Utility for Assisting HDLM Transitions Item 6. Explanation OldHDevName The logical device file name (HDevName) of an HDLM device in HDLM 04-00 is displayed.
The dlmsetboot Utility for Assisting Configuration of an HDLM Boot Disk Environment This utility automates part of the tasks for configuring a boot disk environment in Solaris 9 and Solaris 10. Note that this utility cannot be used in an environment in which cluster software and a volume manager are used. Format /opt/DynamicLinkManager/bin/dlmsetboot -d post-migration-boot-diskdevice-name Parameters -d post-migration-boot-disk-device-name Specifies the HDLM device name of the post-migration boot disk.
configuration definition file (/kernel/drv/dlmfdrv.conf) and the HDLM nexus driver configuration definition file (/kernel/drv/dlmndrv.conf). The sd or ssd device entries are sorted in ascending order of the LDEV information (for example, 9970/9980.15001.066A). If an existing HDLM driver configuration definition file (/kernel/drv/ dlmfdrv.conf) is available, the existing definitions are inherited. If an existing HDLM nexus driver configuration definition file (/kernel/drv/ dlmndrv.
Parameters -d output-directory-name Specifies the name of the directory to which the configuration definition file created by the dlmsetconf utility is to be output. This directory must be specified as an absolute path. You cannot specify multi-byte characters. If output-directory-name is omitted, the configuration definition file will be created in the /kernel/drv directory. Create the output directory before you execute the utility.
LU is managed by HDLM is updated, and applied in the system when the reconfiguration option is specified to restart the host. When the -u parameter is specified, the entry is not added or deleted even though the system configuration is changed (such as by adding or deleting LUs).
The -d parameter is specified: A configuration definition file is created in the /tmp directory: # mkdir /tmp/dlmdir # /opt/DynamicLinkManager/bin/dlmsetconf -d /tmp/dlmdir Example 3 The -r parameter is specified: A new configuration definition file is created in the /kernel/drv directory. Existing information is not inherited.
dlmfdrv.unconf file exists either in the /opt/DynamicLinkManager/ config directory or in a user-defined directory, the file is renamed dlmfdrv.unconf.bak and saved as the previous generation of the definition file. 7-28 • Even when the -d parameter is specified, existing definition information is read from the /kernel/drv/dlmfdrv.conf and /kernel/drv/ dlmndrv.conf files, rather than from the dlmfdrv.conf and dlmndrv.conf files in the specified directory.
• If you want HDLM to manage a device that is not currently managing, comment out the LDEV line (add a hash mark (#) at the beginning of the line) of the definitions for the desired LU in the HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/config/ dlmfdrv.unconf), specify the reconfiguration option, and then restart the host.
The dlmstart Utility for Configuring HDLM Devices This utility configures HDLM devices without needing to restart the host immediately after HDLM is installed. The dlmstart utility for configuring HDLM devices cannot be executed if HDLM devices are being used. The dlmstart utility is valid in Solaris 10 or Solaris 11. Format /opt/DynamicLinkManager/bin/dlmstart [-t | -h ] Parameters -t Uses the existing HDLM driver configuration definition file (/kernel/drv/ dlmfdrv.conf) to configure HDLM devices.
This example uses the existing dlmfdrv.conf content to configure the HDLM devices. # /opt/DynamicLinkManager/bin/dlmstart -t Note • To reconfigure HDLM devices from HDLM version 7.6.0 or earlier, execute the dlmsetconf utility and then restart the host. If you execute the dlmstart utility, an error occurs. If this happens, restart the host to reconfigure the HDLM devices. • If VxVM is installed on the host, do not execute the dlmstart utility.
Format /opt/DynamicLinkManager/bin/dlmvxexclude [-d output-directory-name| -h ] Parameters -d output-directory-name Specifies the path name of the directory to which the file is to be output. You cannot specify multi-byte characters. If the directory specified here contains a regular file having the same name as vxvm_exclude.txt, the confirmation prompt appears. Entering Y or y creates a backup file with the name vxvm_exclude.txt.bak and creates a new vxvm_exclude.txt file.
physical paths. Thus, executing the dlmvxexclude utility outputs entries for HDLM devices that have not actually been created. Entries for these devices are output for each LU, and are sorted in alphabetic order, using the logical device file name as a key.
Used) on page 3-55, Performing a New Installation of HDLM (When Solaris Cluster Is Being Used) on page 3-62, or Performing an Upgrade Installation or Re-installation of HDLM on page 3-73. If you use the -r or -t option when executing this utility, you need to reconfigure the HDLM devices by using the dlmstart utility.
[INSTALLATION_SETTINGS] section This section defines operation information for the installhdlm utility. The [INSTALLATION_SETTINGS] section must be specified first in the installationinformation settings file. However, this section may be preceded by a blank line or comment line. The following table describes the keys to be defined in the [INSTALLATION_SETTINGS] section.
Required or optional? Key name#1 driver_config Explanation#2 Specify whether to configure the HDLM driver by specifying either of the following values: Maximu For m length upgrade For new (characte installatio installati rs) n or reon installatio n Optional Optional 1 Optional Optional 1 Specify whether to use the Optional dlmstart utility to configure HDLM by specifying either of the following values: Optional 1 y: Configure the HDLM driver (default). n: Do not configure the HDLM driver.
#1 Only one key name and value can be specified on each line. #2 Specifying an invalid value causes an error. #3 When you perform a new installation of HDLM or when you upgrade HDLM after the license has expired, make sure that you have a valid license key file. [ENVIRONMENT_SETTINGS] section This section defines operation information for the HDLM command's set operation. This section can be omitted.
Required or optional? Key name#1 Explanation#2 Required Maximum for length Required upgrade (characte for new installati rs) installatio on or ren installati on lio: The Least I/Os algorithm exlio: The Extended Least I/Os algorithm (default) lbk: The Least Blocks algorithm exlbk: The Extended Least Blocks algorithm load_balance_same Specify the number of times _path_use_times the same path can be used for I/O operations when the Round Robin (rr), Least I/Os (lio), or Least Blocks (lbk) algorithm is used
Required or optional? Key name#1 Explanation#2 Required Maximum for length Required upgrade (characte for new installati rs) installatio on or ren installati on Optional 2 Optional Optional 1 Optional Optional 5 trace_file_number Specify the maximum number Optional of trace files (hdlmtr[1-64].log). Optional 2 Optional Optional 3 Optional Optional 4 Optional Optional 3 error_log_number Specify the maximum number Optional of error log files (dlmmgr[1-16].log).
Required or optional? Key name#1 Explanation#2 Required Maximum for length Required upgrade (characte for new installati rs) installatio on or ren installati on off: Disable automatic failback. auto_failback_int Specify the path status check Optional erval interval, which is the interval between the end of the previous check and the start of the next check, in minutes.
Required or optional? Key name#1 Explanation#2 Required Maximum for length Required upgrade (characte for new installati rs) installatio on or ren installati on about the switching of controllers performed by the storage system.#5 You can set the checking interval from 1 to 1440. The default is 10. Legend: Optional: If you omit the key name and value, the following value is used: ¢ ¢ For a new installation The default is used.
#5 The checking interval can be set regardless of whether the dynamic I/O path control function is enabled or disabled.
Note • If VxVM is installed on the host, do not use the -r option or the -t option. Perform the installation without specifying any options.
installux.sh Utility for HDLM Common Installer This utility determines what OS HDLM will be installed on, and installs the corresponding version of HDLM from the DVD-ROM. This utility can also perform unattended installations via a parameter specification.
Log file The installux.sh utility outputs execution information to the log file /var/tmp/hdlm_installux_sh.log. If the hdlm_installux_sh.log log file already exists, the utility appends execution information onto the end of the log file. If the size of the log file is equal to or greater than 1 MB when the utility is executed, the log file is renamed to hdlm_installux_sh2.log, and a new log file is created and named hdlm_installux_sh.log, to which the execution information is output.
# reboot -- -r At the ok prompt, execute the following: boot -r • If HDLM devices are used in ZFS or in SVM, do not use the -r option or the -t option. Perform the installation without specifying any options. After the installation has finished, perform the following steps to configure the HDLM device: See the procedure outlined in Changing the Configuration by Restarting the Host on page 4-37 of Configuration Changes Such as Deleting a Logical Unit on page 4-36.
8 Messages This chapter describes the format and meaning of the message IDs, and also the terms used in the messages and message explanations. For details on the meaning of the return codes output by HDLM when it receives a request from Global Link Manager and measures to take for them, see Return Codes for Hitachi Command Suite Common Agent Component on page 8-136.
□ Return Codes for Hitachi Command Suite Common Agent Component 8-2 Messages Hitachi Dynamic Link Manager (for Solaris) User Guide
Before Viewing the List of Messages This section explains the following information that is needed to locate messages and understand the explanations in the sections from KAPL01001 to KAPL02000 on page 8-4KAPL01001 to KAPL02000. • Format and meaning of the message IDs • Terms used in the messages and message explanations This information is explained below. Format and Meaning of Message IDs Each message has a message ID. The following table shows the format and meaning of message IDs.
• HDLM management target KAPL01001 to KAPL02000 Message ID KAPL01001 -I Message Text The HDLM command completed normally. Operation name = aa...aa, completion time = bb...bb Explanation Details The HDLM command completed successfully. When the view -path, view -lu, or view -drv operation is executed, view(-pstv) is displayed if the Physical Storage View is disabled, and view(-vstv) is displayed if the Physical Storage View is disabled. aa...aa: Specified operation name bb...
Message ID Message Text Explanation retry. For details on the help operation, see help (Displays the Operation Format) on page 6-4. KAPL01006 -W A necessary parameter is not specified. Operation name = aa...aa Details The specified operation does not contain the necessary parameter. aa...aa: Specified operation name Action Execute help operation-name of the HDLM command (dlnkmgr) to check the parameter. Specify the correct parameter, and then retry.
Message ID Message Text Explanation Execute the view operation of the HDLM command (dlnkmgr) to check whether the HDLM manager has started. Start the HDLM manager if it has not started, and then retry the HDLM command. For details on the view operation, see view (Displays Information) on page 6-34. KAPL01013 -E An error occurred in internal Details processing of the HDLM command. An error unrelated to a user Operation name = aa...aa details = operation occurred during bb...bb command processing. aa...
Message ID Message Text Explanation for the relevant HBA port, and then retry. For details on the view operation, seeview (Displays Information) on page 6-34. KAPL01016 -W The target CHA port was not found. Details Operation name = aa...aa The path ID indicated by -pathid and required by the -cha parameter is not an object of HDLM management. aa...
Message ID Message Text Explanation details on the view operation, seeview (Displays Information) on page 6-34. KAPL01019 -W The target path was not found. Operation name = aa...aa Details aa...aa: offline, online, monitor, or view ¢ offline/online operation The specified path does not exist. ¢ view/monitor operation The paths have not been configured because creation of the HDLM environment or configuration changes to the HDLM operating environment have not finished.
Message ID Message Text Explanation Terminate unnecessary applications to increase the amount of free memory, and then retry. KAPL01023 -W The last Online path for the device cannot be placed Offline(C). Details The path specified for the offline operation cannot be set to Offline(C) because it is the last, available path to the LU from the host. Action Use the view operation of the HDLM command (dlnkmgr) to check the status of the paths.
Message ID Message Text Explanation If you want to continue processing of the online operation of the HDLM command for other paths, enter y. If you want to terminate processing, enter n. For details on the online operation, see online (Places Paths Online) on page 6-12. KAPL01040 -W The entered value is invalid. Reenter [y/n]: Details A value other than y and n was entered. Enter y or n. Action Enter y or n. KAPL01041 -E The entered value is invalid. The operation stops. Operation name = aa...
Message ID Message Text Explanation The specified operation is not an operation of the HDLM command. aa...aa: Specified operation name Action Use the help operation of the HDLM command (dlnkmgr) to check the operation name. And then retry. For details on the help operation, see help (Displays the Operation Format) on page 6-4. KAPL01049 -I Would you like to execute the operation? Operation name = aa...aa [y/n]: Details The clear/set operation will be started. To continue the operation, enter y.
Message ID Message Text Explanation If you want to execute online processing, enter y. If you want to terminate processing, enter n. Before you execute the processing, be sure to execute the view operation of the HDLM command (dlnkmgr) to check the path status. For details on the view operation, see view (Displays Information) on page 6-34. For details on the online operation, see online (Places Paths Online) on page 6-12.
Message ID Message Text Explanation If you want to continue processing the offline operation of the HDLM command for other paths, enter y. If you want to terminate processing, enter n. For details on the offline operation, see offline (Places Paths Offline) on page 6-6. KAPL01055 -I All the paths which pass the specified aa...aa will be changed to the Offline(C) status. Is this OK? [y/n]: Details Multiple paths will be collectively set to Offline(C) because the hba or -cha parameter was specified.
Message ID Message Text Explanation enter y. If you want to terminate processing, enter n. KAPL01058 -W The specified parameter value is not needed. Operation name = aa...aa, parameter = bb...bb, parameter value = cc...cc Details A parameter value was specified in a parameter that does not need a parameter value. aa...aa: Specified operation name bb...bb: Parameter name cc...
Message ID Message Text Explanation aa...aa: Number of paths where the online/offline operation was successful (decimal (base-10) number) bb...bb: Online, Online(S), Online(D) or Offline(C) cc...cc: Number of paths where the online/offline operation was unsuccessful (decimal (base-10) number) dd...dd: online or offline Action None. For details on the online operation, see online (Places Paths Online) on page 6-12. For details on the offline operation, see offline (Places Paths Offline) on page 6-6.
Message ID Message Text Explanation None. KAPL01069 -W The entered license key is invalid. Details The entered license key is invalid. Action Enter a valid license key. KAPL01070 -E The entered license key is invalid. Renewal of the license key will now stop. Details The license key renewal processing will be aborted because an invalid license key was entered three times in a row. Action Obtain a valid license key, and then retry. KAPL01071 -I The permanent license was installed.
Message ID Message Text Explanation Re-install HDLM. KAPL01076 -I The permanent license has been installed. Details You need not install a license because a permanent license has already been installed. Action None. KAPL01078 -W KAPL01079 -W The operation terminated because the path configuration changed during execution of the HDLM command. Operation name = aa...aa Details The intermittent error monitoring function cannot be set up because automatic failback is disabled. Details aa...
Message ID KAPL01082 -E Message Text There is no installable license key in the license key file. File name = aa...aa Explanation Details There is no useable license key for HDLM in the license key file. aa...aa: /var/tmp/hdlm_license Action Make sure that the license key file is correct, and then reexecute. /var/tmp/hdlm_license KAPL01083 -I There is no license key file. File name = aa...aa Details There is no license key file in the designated directory: aa...
Message ID KAPL01089 -E Message Text One of the following was executed at the same time as an HDLM command set -lic operation: another set -lic operation, or an update of the license for an update installation. Explanation Action Check the license by using the HDLM command's view -sys lic operation. If necessary, reexecute the HDLM command's set -lic operation. If the same error message is output again, contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM.
Message ID Message Text Explanation to be put online has been reserved by another path. aa...aa: Path ID bb...bb: Code showing reason for error Action Use the format command to confirm that the LU for the path to be put online is not reserved by another host. If the LU is incorrectly reserved, remove the reservation, and then re-execute the online operation for the HDLM command.
Message ID Message Text Explanation HDLM error information to collect the error information, and then contact your HDLM vendor or maintenance company if you have a maintenance contract. KAPL01095 -E An attempt to acquire the HDLM version information has failed. details = aa...aa Details aa...aa: Code showing the reason for the error Action Re-execute the command.
Message ID Message Text Explanation Action None. KAPL01101 -W The target HBA port was not found. Details Operation name = aa...aa The HBA having the HBA_ID specified in the -hbaid parameter could not be found. aa...aa: monitor, offline, online Action Use the view operation of the HDLM command (dlnkmgr view -hba) to check the target HBA port and the HBA_ID of the target HBA port. After that, specify the appropriate HBA_ID, and then retry. KAPL01102 -I All the paths which pass the specified aa...
Message ID Message Text Explanation or -chaid parameter was specified. To collectively place multiple paths Online, enter y. To not collectively place them Online, enter n. aa...aa: CHA, HBA Action If you want to execute the Online processing for the paths which pass the specified target, enter y. If you want to terminate the processing, enter n. KAPL01107 -I The load balancing type specified for individual LUs will become invalid when this operation is executed.
Message ID KAPL01115 -W Message Text The audit log configuration file cannot be opened. Execute the "dlnkmgr view -sys -audlog" command and check whether a normal result is displayed. Explanation Details The audit log configuration file cannot be opened. Action If the "dlnkmgr view -sys audlog" command does not display a normal result, contact your HDLM vendor or maintenance company if there is a maintenance contract for HDLM. KAPL01116 -W The audit log configuration file is Details invalid.
Message ID KAPL01120 -W Message Text A storage system model ID could not be displayed. Details = aa...aa, bb...bb Explanation Details A storage system model ID could not be displayed. aa...aa: Storage recognition information bb...bb: Error code Action Execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM.
Message ID KAPL01134 -I Message Text The target paths are already Online or Online(S). Explanation Details The specified paths are already in the Online or Online(S) status as a result of an online operation. Action Check path status by using the view operation.. For details on the view operation, see view (Displays Information) on page 6-34. For Online(S) paths: To change the status of a path from Online(S) to Online, reexecute the HDLM command using the -hapath parameter.
Message ID Message Text Explanation Use the view operation of the HDLM command (dlnkmgr) to check the setting for the host, storage, and LUs. For details on the view operation, see view (Displays Information) on page 6-34. KAPL01158 -E Dynamic I/O path control cannot be Action applied to the specified storage. Use the view operation of the HDLM command (dlnkmgr) to check the path ID. For details on the view operation, see view (Displays Information) on page 6-34. KAPL01159 -I Paths were added.
Message ID KAPL01162 -I Message Text A path was added. (path ID = aa...aa, storage = bb...bb, iLU = cc...cc) Explanation Details This message displays information about a path added as the result of an add operation. aa...aa: path ID of the added path bb...bb: storage (vendor ID.product ID.serial number) to which the added path is connected cc...cc: LU number to which the added path is connected Action Use the view operation of the HDLM command (dlnkmgr) to verify information about the added paths.
Message ID Message Text Explanation To set all paths to Online or Online(D), enter y. To cancel the operation, enter n. Before you execute the processing, you must check the path status by executing the view operation of the HDLM command dlnkmgr. KAPL01168 -I All P-VOL paths that are connected Details to the LU that has the specified All paths for each specified LU will path ID will be set to Online(D). Is be set to Online or Online(D). this OK? [y/n]: To continue, enter y.
Message ID KAPL01171 -I Message Text The target paths are already aa...aa or bb...bb. Explanation Details The specified paths are already in the Online, Online(S), or Online(D) status as a result of an online operation. aa...aa: Online or Online(S) bb...bb: Online(D) or Online(S)/Online(D) Action Check path status by using the view operation.
Message ID Message Text Explanation aa...aa: virt Action When specifying virtual storage information as a display item, enable the Physical Storage View. KAPL01175 -W If the Physical Storage View is Details enabled, the parameter value If the Physical Storage View is aa...aa cannot be specified for the enabled, the parameter value item parameter. shown cannot be specified. aa...
Message ID Message Text Explanation aa...aa: API trace error information Action Execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL03006 -E An access to the HDLM driver causes an error. (aa...aa) Details This information is required for resolving problems. aa...
Message ID Message Text Explanation This information is required for resolving problems. aa...aa: API trace error information Action Execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL03999 -E An unexpected error occurred.
Message ID Message Text Explanation your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL04003 The startup parameter is invalid. -E Details The HDLM manager internal parameter is invalid.
Message ID Message Text Explanation Use the dlnkmgr set operation to return options settings back to where you had them. If the KAPL04033-W message is not output, restart HDLM manager. If the same error occurs, reinstall HDLM.For details on the view operation, see view (Displays Information) on page 6-34. For details on the set operation, see set (Sets Up the Operating Environment) on page 6-17. KAPL04010 Could not open the error log file.
Message ID Message Text Explanation a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL04013 Input is impossible via the -E communication pipe. RC = aa...aa Details Data could not be read from the pipe file while communicating with an HDLM command. aa...
Message ID Message Text Explanation Execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL04021 HDLM manager information - aa...aa Details -I This information is required for resolving problems. aa...aa: HDLM manager trace information Action None.
Message ID Message Text Explanation HDLM Error Information on page 7-3. KAPL04024 A critical error occurred in the HDLM Details -C manager. (aa...aa) This information is required for resolving problems. aa...aa: HDLM manager trace error information Action Execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM.
Message ID Message Text Explanation KAPL04029 The emergency license expired. -E Action KAPL04030 The temporary license has already -E expired. Action KAPL04031 The emergency license has already -E expired. Action KAPL04032 A fatal error occurred in HDLM. The -C system environment is invalid Details Install a permanent license. Install a permanent license. Install a permanent license. A part of the HDLM configuration file is missing. Action Re-install HDLM.
Message ID Message Text KAPL04036 The path health check for the path -I aa...aa was executed. Number of error paths = bb...bb Explanation Details aa...aa: Number of paths targeted for path health checking. bb...bb: Number of paths determined to have an error by path health checking. Action None. KAPL04037 The path health check completed -I normally. Path ID = aa...aa Details All the paths that path health checking examined are fine. aa...aa: PathID for the path examined by path health checking.
Message ID Message Text KAPL04054 The owner controller will now be -I revised. (number of LUs = aa...aa) Explanation Details aa...aa: number of targeted LUs Action None. KAPL04055 The owner controller was revised. -I (number of changed LUs = aa...aa) Details aa...aa: number of changed LUs Action None. KAPL04056 The owner controller cannot be -W revised because no paths are in the Online status. (LU = aa...aa) Details aa...
Message ID Message Text Explanation not started or contains an error, execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL05011 Could not attach the HDLM driver -E (filter component) to Disk (aa...aa), Partition (bb...bb). (cc...cc:dd...
Message ID Message Text the DLMgetras utility to collect error information, and then contact your vendor or maintenance company. Refer to the HDLM User's Guide for instructions how to execute the DLMgetras utility. KAPL05019 The FO processing in the path -I (aa...aa) finished. The I/O request was processed in the path (bb...bb). Explanation bb...bb: Error code (hexadecimal number) cc...cc: 0 (fixed) Action The I/O being processed is discarded.
Message ID Message Text instructions how to execute the DLMgetras utility. KAPL05039 Received the IOCTL error from Disk -W Driver. errno = aa...aa (bb...bb:cc...cc) Check whether the path is correctly connected to the disk. Explanation Check the message of the HDLM command (dlnkmgr) or the HDLM manager, and then take the appropriate action.
Message ID Message Text Explanation KAPL05041 The contents of the configuration Details -E definition file (dlmfdrv.conf) are A definition in the configuration invalid. (Instance = aa...aa) Execute definition file for the HDLM filter the dlmsetconf command, and then driver is invalid. reboot. (bb...bb:cc...cc:dd...dd) aa...aa:HDLM filter driver instance number bb...bb: Code indicating the error type cc...cc: Error information 1 dd...
Message ID Message Text Explanation Contact the HDLM vendor, or the maintenance company if you have a maintenance contract for HDLM. KAPL05412 An attempt to allocate a bus address Details -W for the HDLM device instance aa...aa An attempt to assign a physical failed. The HDLM device could not be path name (device address) to attached. the HDLM filter driver instance that you tried to attach has failed. aa...
Message ID Message Text how to execute the dlmsetconf command. Explanation bb...bb: Instance number (decimal number) Action Check whether the path is correctly connected to the disk. Execute the dlmsetconf utility for creating the HDLM driver configuration definition file, and then reboot the host. KAPL05451 The path (aa...aa) has been deleted. Details -I The dynamic LU deletion function deleted a path connected to the LU that was deleted. aa...aa: Path ID Action None. KAPL05452 The path (aa...
Message ID Message Text Explanation problems, this message is sometimes output in Oracle ASM environments. If this is the case, you can ignore this message because the reported problem has no effect on the system environment. KAPL05515 A path error occurred. The label -W might be invalid. (ErrorCode = aa...aa, PathID = bb...bb) Details A path error occurred. The label might be invalid. aa...aa: ErrorCode bb...bb: PathID Action Make sure that the path is connected.
Message ID Message Text Refer to the HDLM User's Guide for instructions how to execute the DLMgetras utility. Explanation name of the HDLM device to the filter driver has failed. Action Check whether the HDLM driver has started normally. KAPL05905 No authority to execute the -W dlmsetdevname command. Do not execute the command unless it is in a boot script. Details You do not have a permission required to execute dlmsetdevname. Action Do not execute the command unless it is in a boot script.
Message ID Message Text Explanation bb...bb: Target memory size (hexadecimal number) Action Check whether the HDLM driver has started normally. If it has not started or contains an error, execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM and report the error and detail code.
Message ID Message Text Explanation Action Check the message of the HDLM command (dlnkmgr) or HDLM manager, and then take the appropriate action. If you do not know the appropriate action, execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM and report the error and detail code.
Message ID Message Text KAPL06014 Could not write emergency -E information into the emergency information buffer. (aa...aa:bb...bb) Explanation Details Urgent information from the filter driver was destroyed without being written to the urgent information buffer because the attempt to reserve memory for the information failed. aa...aa: Message code (hexadecimal number) bb...bb: Buffer size (hexadecimal number) Action Check whether any other error occurred.
Message ID Message Text Explanation cc...cc: Detailed information 2 (decimal (base-10) number) dd...dd: Detailed information 3 (decimal (base-10) number) Action None. KAPL07824 -I The owner controller of the LU connected to the path (aa...aa) was changed to (bb...bb). Details The owner controller of the LU connected to the path with the ID shown in the message was changed. aa...aa: Path ID of the changed LU. (same as PathID of view path) (Decimal number) bb...bb: Owner controller ID after the change.
Message ID Message Text Explanation ¢ When a path error is detected by a path health checking or the online operation: Displays 0x000F0000(Fixed). ¢ When a path error is detected through an I/O error: Displays the OS error code. cc...cc: Fixed at 0x00000000 Action Check the path in which the error was detected. KAPL08022 A path error occurred. ErrorCode = -E aa...aa, PathID = bb...bb, PathName = cc...cc.dd...dd.ee...ee.ff...ff, DNum = gg...gg, HDevName = hh...
Message ID Message Text Explanation hh...hh: Host device name (same as HDevName of view path) Action There could be an error in the path. See What To Do for a Path Error on page 5-3 and restore the path displayed in the message to running status. KAPL08023 A path was recovered. PathID = -I aa...aa, PathName = bb...bb.cc...cc.dd...dd.ee...ee, DNum = ff...ff, HDevName = gg...gg Details aa...aa: Path ID (same as PathID of view -path) (decimal (base-10) number) bb...
Message ID Message Text KAPL08027 A path was excluded from the items -E subject to automatic failback. PathID = aa...aa Explanation Details A path was excluded from being subject to automatic failbacks because the system judged that an intermittent error was occurring in that path. aa...aa: Path ID (same as PathID of view -path) (decimal (base-10) number) Action An intermittent error has occurred. Check the path for any possible problems.
KAPL09001 to KAPL10000 Message ID KAPL09002 -E Message Text The disk does not have sufficient free space. Explanation Details The file system of the installation does not have sufficient free capacity to install HDLM. Action Delete unnecessary files to increase the amount of free capacity, and then retry. KAPL09008 -W The license code is invalid. KAPL09009 -E The license code is invalid. The HDLM installation program will now terminate. Action Double-check the license code, and then re-enter it.
Message ID Message Text Explanation Network Objectplaza Trace Library 2 fails again, contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. KAPL09021 -E An attempt to register the bundle PP name of Hitachi Network Objectplaza Trace Library 2 failed. Details An attempt to register the PP name of HNTRLib 2 has failed. Action Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM.
Message ID Message Text Explanation /opt/HDVM/agent/classes/com/ Hitachi/soft/HiCommand/DVM/ agent/module/ HDLMManager.class /opt/HDVM/agent/docroot/ webstart/hdlm /opt/HDVM/agent/docroot/ hdlmhelp /opt/HDVM/agent/classes/jp /opt/HDVM/agent/classes/com/ Hitachi/soft/HiCommand/DVM/ agent/module/hdlm KAPL09026 -I Hitachi Network Objectplaza Trace Library 2 wasn't removed because it was being used for other products.
Message ID Message Text Explanation None. KAPL09077 -I The temporary license was installed. The license expires on aa...aa. Details A temporary license was installed. aa...aa: Year (4 digits)/month (01-12)/day (01-31) Action Install a permanent license by the expiration day. KAPL09078 -I The emergency license was installed. The license expires on aa...aa. Details An emergency license was installed. aa...
Message ID Message Text now stop. Obtain a valid license key, and then re-install HDLM. Explanation license key was entered three times in a row. Action Obtain a valid license key, and then re-install HDLM. KAPL09088 -E The entered license key is invalid. The HDLM installation will now terminate. Obtain a valid license key, and then re-install HDLM. Action KAPL09089 -W License information cannot be acquired. Details Obtain a valid license key, and then re-install HDLM.
Message ID Message Text Explanation Install a permanent license at a later time. KAPL09091 -E A fatal error occurred in HDLM. Details The system environment is invalid. A part of the HDLM configuration Contact your HDLM vendor or the file is missing. maintenance company if there is a Action maintenance contract of HDLM. Contact your HDLM vendor or the maintenance company if there is a maintenance contract of HDLM. KAPL09095 -E The entered value is invalid. The installation has been stopped.
Message ID Message Text Explanation Alternatively, cancel the installation, save the correct license key file in the designated directory, and then re-execute the HDLM installation program. /var/tmp/hdlm_license KAPL09115 -W An attempt to delete the license key file has failed. File name = aa...aa Details An attempt to delete the license key file has failed. aa...aa: /var/tmp/hdlm_license Action If a license key file exists, delete it.
Message ID Message Text Explanation result of a user operation occurred. Action Execute the utility for collecting HDLM error information (DLMgetras), and acquire obstacle information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM and report the error and detail code. KAPL09140 -E HDLM cannot be installed in this zone. Install it in the global zone. Details HDLM cannot be installed in a nonglobal zone. Install HDLM in a global zone.
Message ID Message Text Explanation aa...aa: One of the following file names: ¢ ¢ ¢ ¢ /kernel/drv/dlmfdrv.conf /opt/DynamicLinkManager/ config/dlmfdrv.unconf /kernel/drv/ dlmfdrv.conf.bak /opt/DynamicLinkManager/ config/dlmfdrv.unconf.bak bb...bb: Internal code (decimal number) Action a. Execute the DLMgetras utility for collecting HDLM error information to obtain the error information. Contact your HDLM vendor or your maintenance company if you have a maintenance contract for HDLM.
Message ID KAPL09149 -E Message Text An attempt to update the file has failed. File name = aa...aa, Details = bb...bb. Please contact product support. Explanation Details An HDLM driver configuration definition file or HDLM driver unconfiguration definition file, aa...aa, could not be updated. aa...aa: One of the following file names: ¢ ¢ ¢ ¢ /kernel/drv/dlmfdrv.conf /opt/DynamicLinkManager/ config/dlmfdrv.unconf /kernel/drv/ dlmfdrv.conf.bak /opt/DynamicLinkManager/ config/dlmfdrv.unconf.bak bb...
Message ID Message Text Explanation by HDLM, edit the HDLM driver unconfiguration definition file (/opt/ DynamicLinkManager/ config/dlmfdrv.unconf) and re-execute dlmsetconf utility with -u option. KAPL09179 -I Data for maintenance: aa...aa bb...bb c. When you use VxVM, remove HDLM devices from VxVM. The dlmvxexclude utility assists this. d. Restart the host with the reconfiguration option specified. Details aa...aa: Message output location information (decimal (base-10) number) bb...
Message ID KAPL09190 -W Message Text The installation information settings file is not specified. Explanation Details The installation-information settings file is not specified for the second parameter in the installhdlm utility for installing HDLM. Action Make sure that the parameters in the installhdlm utility are appropriate, and then try again. KAPL09191 -W The installation information settings file does not exist.
Message ID Message Text Explanation Check the absolute path of the license key file, and then re-install HDLM. KAPL09205 -E There is no license key file. The HDLM installation will now terminate. File name = aa...aa Details There is no license key file. The HDLM installation will now terminate. aa...aa: Character string that the user entered Action Check the absolute path of the license key file, and then re-install HDLM. KAPL09206 -W The entered value is invalid.
Message ID Message Text Explanation aa...aa: installhdlm, pkgadd, dlnkmgr, dlmsetconf, or dlmstart Action None. KAPL09212 -E aa...aa ended abnormally. Details aa...aa ended abnormally. aa...aa: installhdlm, pkgadd, dlnkmgr, dlmsetconf, or dlmstart Action Check the error message that was output just before this message, and then perform the action indicated in that error message. KAPL09213 -W An error occurred during aa...aa processing. Details Although the aa...
Message ID Message Text Explanation The /var/tmp directory does not exist. Prepare the /var/tmp directory, and then retry the operation. If the error code is 5: You do not have write permission for the /var/tmp directory. Make sure you have access permissions for the /var/tmp directory. If the error code is 6: You do not have write permission for the file system of the /var/tmp directory. Make sure you have access permissions for the file system of the /var/tmp directory.
Message ID Message Text Explanation Make sure that sufficient free capacity exists for the /var/tmp directory. If the capacity is insufficient, allocate the required amount of capacity, and then retry the operation. For details about the required free capacity, see Disk Capacity Requirements on page 3-36 in Memory and Disk Capacity Requirements on page 3-36. KAPL09217 -E An error occurred during reading of the installation information settings file. Error Code = aa...aa,bb...
Message ID Message Text Explanation permission, re-execute installhdlm. If aa...aa is installhdlm_analysis: The correct location of installhdlm_analysis is the same directory as installhdlm. Make sure that the utility you want to execute is in the correct location and that you have the required permissions. If a utility is not in the correct location, copy the required files to the correct location, and then retry the operation.
Message ID Message Text Explanation Specify the revised installation information settings file, and then retry the operation. KAPL09221 -W The definition of the installation information settings file includes an unusable character. Error Code = aa...aa,bb...bb, line = cc...cc Details A character that cannot be used in a non-comment line was used. aa...aa: Error number (decimal number) that specifies the executed processing bb...bb: Return value (decimal number) of the executed processing cc...
Message ID Message Text Explanation Specify the revised installation information settings file and then retry the operation. KAPL09229 -W The definition of the installation information settings file includes an invalid section name. Error Code = aa...aa,bb...bb, line = cc...cc Details A section that does not exist is included. aa...aa: Error number (decimal number) that specifies the executed processing bb...bb: Return value (decimal number) of the executed processing cc...
Message ID Message Text Explanation Specify the revised installation information settings file and then retry the operation. KAPL09232 -W The composition of the definition of the installation information settings file is invalid. Error Code = aa...aa,bb...bb, line = cc...cc Details A key, key value, or equal sign (=) is missing. aa...aa: Error number (decimal number) that specifies the executed processing bb...bb: Return value (decimal number) of the executed processing cc...
Message ID KAPL09235 -E Message Text The log file cannot be output to its destination because the environment is invalid. Error Code = aa...aa Explanation Details The log file cannot be output in this environment, because certain environment settings are invalid. aa...aa: Error number (decimal number) that specifies the executed processing. Action If the error code is 1: The log file destination directory does not exist. Prepare the directory, and then retry the operation.
Message ID Message Text Explanation exists. If the directory exists, delete it by using the following command: # rm -rf /var/opt/ DynamicLinkManager However, if you interrupted an upgrade installation or reinstallation, do not delete this directory. Also, execute the following command to check the status. # pkginfo -l DLManager If partially installed is displayed for the status, perform an update installation. KAPL09238 -W The specified file is not a normal one. Fail name = aa...
Message ID Message Text Explanation installation utility (dlminstcomp). If the error code is 5: Install the prerequisite JDK, and then execute the HDLM component installation utility (dlminstcomp). If the error code is other than the above: Contact your HDLM vendor or your maintenance company if you have a maintenance contract for HDLM. KAPL09242 -E No usable JDK or JRE exists. Details The prerequisite JDK is not installed on the system.
Message ID Message Text Explanation Contact your HDLM vendor or your maintenance company if you have a maintenance contract for HDLM. KAPL09248 -W A parameter is invalid. parameter = aa...aa Details An invalid parameter was specified. aa...aa: Specified parameter (character string) Action Execute the HDLM component installation utility (dlminstcomp) with the -h parameter specified to check the utility format, and then retry the operation. KAPL09288 -E An error occurred during "aa...aa" processing.
Message ID Message Text Explanation removehdlm: DVD-ROM-mountpoint/HDLM_Solaris/IPS/ remove_scripts/removehdlm KAPL09290 -I A user operation ended removehdlm. Details The process has been terminated by an operation such as Ctrl+C. Action Re-execute removehdlm. KAPL09291 -E This version of the OS is not supported. Please remove HDLM on a supported OS. Details This version of the OS is not supported. Action Execute onto the OS which can be supported by HDLM.
Message ID Message Text Explanation An SP cannot be applied because HDLM is not installed on the system. Action Check whether HDLM has been correctly installed. KAPL09504 -E The language environments of HDLM and the Service Pack are different. Details The Japanese SP was applied to the English edition of HDLM, or vice versa. Action Acquire the SP that has the same language as the installed HDLM, and then try again. KAPL09505 -E aa...aa cannot be applied to the installed bb...bb.
Message ID Message Text Explanation Installation of SP (aa...aa) is stopped. aa...aa: The version of the SP to be installed bb...bb: The version of the installed HDLM or the version of the installed SP Action You do not have to install the SP. Keep using HDLM as is. KAPL09601 -E Cannot install in this system. Install HDLM on a supported OS. Details HDLM cannot be installed on this system. Action Execute the installation in a supported OS.
Message ID Message Text Explanation ¢ KAPL09604 -E A system error occurred. The installation will now stop. (command = aa...aa, error code = bb...bb) Confirm that the OS command exists, and that you have the proper permission. Details aa...aa: Command that failed bb...bb: Return value Action Acquire the log /var/tmp/ hdlm_installux_sh.log, and then contact your HDLM vendor or, if you have a maintenance contract, contact the maintenance company.
Message ID Message Text Explanation The first parameter must be a directory to which collected information will be output. aa...aa: First parameter Action Check the parameters of the DLMgetras utility for collecting HDLM error information, and then retry. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL10004 The parameter contains an -W incorrect value. Value = aa...aa Details The first parameter must be a directory.
Message ID Message Text Explanation Action Check whether the specified file for defining the information to be collected exists, and check whether you have access permission for the specified file. KAPL10007 A directory has been specified in -W the third parameter. Value = aa...aa Details The -f parameter is specified to a directory. aa...aa: Third parameter Action Check the parameters of the DLMgetras utility for collecting HDLM error information, and then retry.
Message ID Message Text Explanation DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL10010 A root directory has been specified. Details -W Line = aa...aa The root "/" has been specified as a directory to be collected in the file for defining the information to be collected. aa...aa: Line number of the file for defining information to be collected (decimal number) Action Delete the coding of the root directory from the specified file.
Message ID Message Text KAPL10012 The specified file or directory does -W not exist. Line = aa...aa, Value = bb...bb Explanation Details The specified file or directory does not exist in the file for defining the where information is to be collected. aa...aa: Line number of the file for defining which information to be collected (decimal (base-10) number) bb...
Message ID Message Text Explanation DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL10014 You lack read permission for the -W specified directory. Line = aa...aa, Value = bb...bb Details You lack read permission for the specified directory in the file for defining information to be collected. aa...aa: Line number of the file for defining information to be collected (decimal number) bb...
Message ID Message Text Explanation A root "/" cannot be specified in a directory to which collected information is output. Action Check the parameters of the DLMgetras utility for collecting HDLM error information, and then re-execute. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL10017 You lack privileges for executing -W the utility for collecting HDLM error information.
Message ID Message Text Explanation None. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL10030 A user terminated the utility for -I collecting HDLM error information. Details Processing of the DLMgetras utility for collecting HDLM error information has been terminated because the user replied to the confirmation with an n response. Action None.
Message ID Message Text Explanation bb...bb: cq command output message Action An error occurred during copying of the file to be collected. The user environment might have been unstable. Check the system configuration. KAPL10035 An attempt to archive the error Details -E information failed. Details = aa...aa Execution of the tar command failed. aa...aa: tar command output message Action See the details in the message, and then remove the cause of the error.
Message ID Message Text Explanation None. KAPL10050 The utility for collecting HDLM error Details -I information started. Start time = The utility for collecting HDLM aa...aa (GMT aa...aa) error information started. aa...aa: year (4 digits)/month/ date hour:minute:second Action None. KAPL10201 usage: dlmsetconf [-d output-W directory-name] [-c|-r|-u|-t] Details Displays how to use dlmsetconf. Action Specify appropriate parameters, and then retry. KAPL10202 A backup file (aa...
Message ID Message Text Explanation aa...aa: Name of the configuration definition file Action Check the system configuration. KAPL10223 An error occurred when releasing -E the memory. Details An attempt to release memory has failed. Action Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM if this error occurs a number of times. KAPL10224 An error occurred when creating -E the configuration definition information.
Message ID Message Text KAPL10229 Internal processing error. -E Explanation Details An internal processing error occurred. Action Internal processing error. Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. KAPL10230 The aa...aa directory could not be -E opened. Errno = bb...bb (cc...cc) Details An attempt to open a directory has failed. aa...aa: Directory name bb...bb: Errno (decimal number) cc...
Message ID Message Text Explanation Remove the error by referencing Errno, and then retry. KAPL10236 Could not allocate memory. Errno -E = aa...aa (bb...bb) Details An attempt to reserve memory has failed. aa...aa: Errno (decimal number) bb...bb: Errno string Action Allocate the necessary amount of memory by freeing up unnecessary resources. KAPL10237 The property information (aa...aa) -E on a device (bb...bb) could not be collected.
Message ID Message Text Explanation Action Remove the error by referencing Errno, and then retry. KAPL10242 To configure HDLM, device files of -I sd/ssd devices managed by HDLM are removed after reboot. Do you want to continue? [y/n] Details This message is confirming execution of the dlmsetconf utility for creating the HDLM driver configuration definition file. To configure HDLM, the logical device files of the HDLM-managed sd or ssd device are deleted at reboot.
Message ID Message Text Explanation Action Check the system configuration. KAPL10251 An attempt to retrieve information Details -W for the directory aa...aa failed. An attempt to collect directory Function = readdir, errno = bb...bb information has failed. (cc...cc) aa...aa: Directory name bb...bb: Errno (decimal number) cc...cc: Errno string Action See errno to eliminate the error, and then retry. KAPL10252 An attempt to read the contents of -W symbolic link aa...aa failed.
Message ID Message Text Explanation Processing will stop because an incorrect value was entered three or more times for the message confirming the execution of the dlmsetconf utility for creating the HDLM driver configuration definition file. Action When executing the utility, enter the correct value. KAPL10256 The user terminated the operation. -I Details The user has interrupted processing. Action None. KAPL10257 The output file (aa...aa) could not -W be opened. Errno = bb...bb (cc...
Message ID Message Text Explanation properly connected to the server. If there are any errors, remove the error, and then re-execute the command. If the disk is connected properly, contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. KAPL10259 No authority to execute the -W dlmsetconf utility. Details You do not have a permission required to execute the dlmsetconf utility. Action Retry as a user with root permissions.
Message ID Message Text Explanation command differs from the output of the dlnkmgr view -path command. Action Check the input file, and then retry. For details on the dlmlisthdev utility, see The dlmlisthdev Utility for Assisting HDLM Transitions on page 7-20. KAPL10316 Could not allocate memory. Size = -E aa...aa. Execute the DLMgetras utility to collect error information, and then contact your vendor or maintenance company. Details Securing of memory failed. aa...
Message ID Message Text Explanation Confirm the existence of the file and access permission. If there is no problem, contact your HDLM vendor or maintenance company if you have a maintenance contract for HDLM. KAPL10334 Ioctl aa...aa failed with errno = -E bb...bb. Execute the DLMgetras utility to collect error information, and then contact your vendor or maintenance company. Details An attempt to issue an IOCTL call has failed. aa...aa: IOCTL Code bb...
Message ID Message Text KAPL10371 The HDLM support disk information -W does not exist in the configuration file (dlmfdrv.conf). Check the system configuration, and then retry the command. Explanation Details The /kernel/drv/dlmfdrv.conf file does not contain the data of the HDLM-managed devices. Action Check the system configuration, and then retry. KAPL10372 An error occurred when collecting -E the catalog message. ErrorCode = aa...aa(bb...
Message ID Message Text Explanation dlmcfgmgr Utility for Managing the HDLM Configuration on page 7-17. KAPL10376 The device (aa...aa) could not be -E opened. ErrorCode = bb...bb (cc...cc) Details An attempt to open the device managing the HDLM filter driver has failed. aa...aa: Name of the device managing the HDLM filter driver bb...bb: Errno (decimal number) cc...cc: Errno string Action Take action in accordance with ErrorCode (errno), and then retry.
Message ID Message Text Explanation dlmcfgmgr Utility for Managing the HDLM Configuration on page 7-17. KAPL10382 There is no instance of the HDLM Details -W nexus driver that manages the path This LU was not added because of the LU. This LU (LDEV = aa...aa) the HDLM nexus driver instance is not added. FilterPath = bb...bb. does not exist in the LU. If you want to use the LU, reboot aa...aa: LDEV name the host. bb...bb: Filter Path Action If you want to use the LU, reboot the host. KAPL10383 The text aa.
Message ID Message Text the LDEV name has changed, reboot the host. KAPL10387 The path of the another LU (LDEV -W = aa...aa) managed by HDLM has the same value as the property (bb...bb). The configuration of the LU (LDEV = cc...cc) will not be changed. If you want to use the LU, reboot the host. Explanation Action If the path configuration of the LU shown by the LDEV name has changed, reboot the host. If there is no change to the configuration, do nothing.
Message ID Message Text KAPL10391 The output file (aa...aa) could not -E be opened. ErrorCode = bb...bb (cc...cc). Resolve the error by referencing ErrorCode. Explanation Details An attempt to open the output file has failed. aa...aa: Output file name bb...bb: Errno (decimal number) cc...cc: Errno string Action Remove the error by referencing ErrorCode (errno). KAPL10392 The catalog file (aa...aa) could not -E be opened. ErrorCode = bb...bb (cc...
Message ID Message Text to complete execution and then run the command. Explanation cannot be simultaneously executed for multiple processes. Action Wait until the dlmcfgmgr utility ends, and then re-execute the utility. For details on the dlmcfgmgr utility, see The dlmcfgmgr Utility for Managing the HDLM Configuration on page 7-17. KAPL10398 An attempt to acquire the process -E ID has failed. ErrorCode = aa...aa (bb...bb).
Message ID Message Text Explanation Action It is not necessary to execute this command. For details on the dlmvxexclude utility, see The dlmvxexclude Utility for Assisting Creation of the VxVM Configuration File on page 7-31. KAPL10704 dlmvxexclude completed -I successfully. Details The dlmvxexclude utility for assisting creation of the VxVM configuration file completed successfully. Action None.
Message ID Message Text Explanation Creation of the VxVM Configuration File on page 7-31. KAPL10708 No authority to execute -E dlmvxexclude. Details You do not have permission to execute the dlmvxexclude utility for assisting creation of the VxVM configuration file. Action Execute by a user with root permissions. For details on the dlmvxexclude utility, see The dlmvxexclude Utility for Assisting Creation of the VxVM Configuration File on page 7-31.
Message ID Message Text Explanation A valid entry does not exist in the /kernel/drv/dlmfdrv.conf file. Action Check the system configuration and then execute the dlmsetconf utility for creating the HDLM driver configuration definition file. KAPL10713 The aa...aa directory could not be -E opened. Errno = bb...bb (cc...cc) Details An attempt to open a directory has failed. aa...aa: Directory name bb...bb: Errno (decimal number) cc...cc: Errno string Action Check the system configuration and then retry.
Message ID Message Text Explanation aa...aa: Directory name Action Specify the correct directory. KAPL10717 The directory specified by the -d -E option does not exist. Details A non-existing directory was specified. Action Specify the correct directory. KAPL10718 An error occurred when collecting -E the catalog message. Errno = aa...aa (bb...bb) Details A message could not be acquired from the message catalog. aa...aa: Errno bb...
Message ID Message Text Explanation message confirming execution of the dlmvxexclude utility for assisting creation of the VxVM configuration file. Action Retry dlmvxexclude command. For details on the dlmvxexclude utility, see The dlmvxexclude Utility for Assisting Creation of the VxVM Configuration File on page 7-31. KAPL10722 The user terminated the operation. -I Details The execution of the dlmvxexclude utility for assisting creation of the VxVM configuration file was stopped. Action None.
Message ID Message Text Explanation aa...aa: Physical-device file path of the sd or ssd device that corresponds to the HDLM device for which the path was deleted Action None. KAPL10734 The boot disk configuration cannot -W be changed. Details The boot disk configuration cannot be changed. Action Migrate the boot environment either to an internal disk or to another LU, and then change the original boot-disk configuration. KAPL10735 The primary-path information file -E could not be opened.
Message ID Message Text Explanation None. KAPL10752 The dlmsetboot utility completed -I successfully. End time = aa...aa Details The dlmsetboot utility completed successfully. aa...aa: year (4 digits)/month/ date hour:minute:second Action None. KAPL10753 The dlmsetboot utility terminated -E unsuccessfully. End time = aa...aa Details The dlmsetboot utility terminated unsuccessfully. aa...
Message ID Message Text Explanation KAPL10756 The entered value is invalid. The Details -E dlmsetboot utility processing will Processing will stop because an now stop. Execute the utility again. incorrect value was entered three End time = aa...aa or more times for the message confirming execution of the dlmsetboot utility. aa...aa: year (4 digits)/month/ date hour:minute:second Action Re-execute the dlmsetboot utility.
Message ID Message Text Explanation KAPL10760 The post-migration boot disk device Details -E is not an HDLM management The post-migration boot disk target. (aa...aa) device is not an HDLM management target. (aa...aa: Post-migration boot disk device name (cXtYdZ)) Action Set the post-migration boot disk device as an HDLM management target. KAPL10761 The file system of the pre-E migration boot disk device is not a UFS file system.
Message ID Message Text Explanation # The following information is output: /usr/sbin/installboot /usr/sbin/newfs /usr/sbin/mount /usr/sbin/ufsdump /usr/sbin/ufsrestore /usr/sbin/eeprom Action After the dlmsetboot utility terminates, check the contents of the messages by the command. If the contents are incorrect, correct them and then try to execute this utility again. KAPL10765 Do you use Sun Cluster or SVM? -I [y/n] Details Do you use Sun Cluster or SVM? [y/n] Action Enter y or n.
Message ID Message Text Explanation KAPL10768 The post-migration boot disk device Details -I configuration was updated. The post-migration boot disk (aa...aa) (bb...bb) device configuration was updated. aa...aa= configuration file name (absolute path)# bb...bb= backup file name of preupdate configuration file (absolute path) # The following information is output: /etc/system rootdev parameter (only specified) /etc/vfstab /etc/dumpadm.conf DUMPADM_DEVICE=... Action None.
Message ID Message Text Explanation aa...aa: Physical device path of the sd or ssd device associated with the HDLM device for which a path was successfully added Action None. KAPL10772 Acquisition of the major number of -E the nexus driver failed. ErrorCode = aa...aa (bb...bb) Details Acquisition of the major number of the nexus driver failed. ErrorCode = aa...aa (bb...bb) aa...aa: Error code bb...
Message ID Message Text Explanation aa...aa: SerialNo bb...bb: iLu Action None. KAPL10776 For LU information that is not in the Details -I storage migration information For LUs that are not listed in the definition file (/opt/ storage migration information DynamicLinkManager/config/ definition file, do you want to take dlmmiginfo.
Message ID Message Text Explanation KAPL10780 The storage migration information Details -E definition file is invalid. It is empty. The storage migration information (aa...aa) definition file is invalid. It is empty. (aa...aa) file name Action Recreate the storage migration information definition file, and then retry dlmsetconf. KAPL10781 The storage migration information -E definition file is invalid. It contains a property without a value. (aa...aa: bb...
Message ID Message Text KAPL10784 The storage migration information -W definition file does not exist. (aa...aa) Explanation Details The storage migration information definition file does not exist. aa...aa: file name Action Create the storage migration information definition file, and then retry dlmsetconf. KAPL10785 An error occurred during creation -E of a file used for confirmation. (aa...aa) Details An error occurred during creation of a file used for confirmation. aa...
Message ID Message Text KAPL10863 The aa...aa service cannot be -E enabled. Explanation Details A service cannot be enabled. aa...aa: Service name Action Check the file /var/opt/ DynamicLinkManager/log/ dlmstart[1-2].log. If you cannot determine why the service cannot be enabled, use DLMgetras to collect maintenance information. After that, contact either your HDLM vendor or, if you have an HDLM maintenance contract, your contracted support company. KAPL10864 The aa...aa driver cannot be -E removed.
Message ID Message Text Explanation None KAPL10868 Log in as the root user before using Details -E the dlmstart utility. Log in as the root user before using the dlmstart utility. Action Execute the command as the root user. KAPL10869 One or more HDLM devices cannot -E be removed because they are in use. Details One or more HDLM devices cannot be removed because they are in use. Action Check whether any HDLM devices are in use.
Message ID Message Text Explanation Execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. For details on the DLMgetras utility, see "7.2 DLMgetras utility for collecting HDLM error information" in the Hitachi Dynamic Link Manager User Guide for Solaris. KAPL10874 dlmstart cannot be executed in this Details -E zone. Execute dlmstart in the dlmstart is not running in the global zone.
Message ID Message Text Explanation Make sure that the other instance of dlmstart finished successfully. KAPL10880 An attempt to generate a link -E failed. The devfsadm command failed. Details devfsadm failed. Action Execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. For details on the DLMgetras utility, see "DLMgetras utility for collecting HDLM error information".
Message ID Message Text Explanation aa...aa: LDEV name bb...bb: Filter path Action If you want to use the LU, reboot the host. KAPL10953 Deletion of the path (aa...aa) -E failed. (Function: bb...bb, ErrorCode = cc...cc (dd...dd)) Details Deletion of a path failed. aa...aa: Filter path name bb...bb: Function name cc...cc: errno dd...dd: errno string Action Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM.
Message ID Message Text Explanation ¢ Get HDLM Alert Driver Status ¢ Get SNMP Trap Information ¢ Set SNMP Trap Information ¢ Set LU Load Balance ¢ Get Path Status Log Information ¢ Get Local Time ¢ Add Path Information ¢ Delete Path Information ¢ Set Storage Identification Information Action None. KAPL11902 aa...aa has started. PathID = -I bb...bb Details The operation has started on the host. aa...aa: Operation (character string) ¢ Online ¢ Offline bb...
Message ID Message Text Explanation ¢ Get Local Time ¢ Add Path Information ¢ Delete Path Information ¢ Set Storage Identification Information Action None. KAPL11904 aa...aa has completed abnormally. -E Error status = bb...bb Details The operation has completed abnormally on the host. aa...
Message ID Message Text KAPL11905 An unexpected error occurred. -E Explanation Details An exception occurred during processing on the host. Action Execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if you have a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL11906 GUI information - aa...
Message ID Message Text Manager, and execute the "dlnkmgr view -sys -audlog" command and check the setting. KAPL13602 The audit log configuration file -W cannot be opened. Execute the "dlnkmgr view -sys -audlog" command and check whether a normal result is displayed. Explanation Action Restart the HDLM Manager, and execute the "dlnkmgr view -sys -audlog" command, and then specify the desired setting by using the "dlnkmgr set audlog" command or the "dlnkmgr set -audfac" command as necessary.
Message ID Message Text Explanation Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. KAPL13606 An error occurred during processing -W to output the audit log configuration file. Details An internal error occurred during output of the audit log data. Action Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM.
Message ID Message Text Explanation KAPL15027 An attempt to output information -E about exclusion from VxVM management has failed. Command Line = aa...aa Details KAPL15060 DLMgetras was invoked. Command -I Line = aa...aa Details KAPL15061 DLMgetras successfully executed. -I Command Line = aa...aa Details KAPL15101 Clear operation was completed -I successfully. Command Line = aa...aa Details aa...aa: Name of the utility that the user executed aa...
Message ID Message Text Explanation KAPL15109 Information about HDLM-I management targets was successfully displayed. Command Line = aa...aa Details KAPL15110 An attempt to display information -E about HDLM-management targets has failed. Command Line = aa...aa Details aa...aa: Command that the user executed aa...aa: Command that the user executed KAPL15111 The HDLM command was started or Details -W stopped by the user who does not aa...aa: Command that the user have the authority.
Message ID Message Text Explanation KAPL15401 HDLM Manager successfully started. -I - KAPL15402 Could not start the HDLM manager. -E - KAPL15403 HDLM Manager successfully stopped. -I - KAPL15404 The HDLM Manager was executed by -W the user who does not have the authority.
Message ID Explanation The target path of the offline operation is the last, remaining path connected to the device and therefore, cannot be placed in the offline status. Action Click Refresh to update the host information, check the path status, and then retry the offline operation. 1015 Details The Offline path cannot be placed Online. Action Remove the error in the path, and then retry. 1016 Details The target path(s) are already Online.
Message ID Explanation you have a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. 1026 Details The acquisition of path information has been aborted, because the path configuration was changed while the system was attempting to acquire the path information. Action Refresh the host information, check the path status, and then perform the operation again.
Message ID Explanation Refresh the host information, and then perform the operation again. If the same error occurs even after removing the reservation, execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or maintenance company if you have a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3.
Message ID Explanation contact your HDLM vendor or maintenance company if there is a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. 1035 Details An attempt to acquire information about the HDLM version or SP version has failed. Action Re-execute the command.
Message ID Explanation Refresh the host information, check the path status, and then perform the operation again. 1045 Details A parameter is invalid. Action Refresh the host information, and then perform the operation again. If the same error occurs even again, execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or maintenance company if you have a maintenance contract for HDLM.
8-142 Messages Hitachi Dynamic Link Manager (for Solaris) User Guide
A Sun Cluster 3.2 Commands This section describes the Sun Cluster 3.2 commands. □ Sun Cluster 3.2 Commands Sun Cluster 3.
Sun Cluster 3.2 Commands The table below shows the correspondence between Sun Cluster 3.1 commands and Sun Cluster 3.2 commands used in this manual. For details about the command options, see the documentation for Sun Cluster 3.2.# # Note that there are no differences between Sun Cluster 3.2 commands and Oracle Solaris Cluster 3.3 commands. Table A-1 Correspondence Between Sun Cluster 3.1 Commands and Sun Cluster 3.2 Commands Commands for Sun Cluster 3.1 Commands for Sun Cluster 3.
Commands for Sun Cluster 3.1 Commands for Sun Cluster 3.2 scswitch -F -D device-group-name cldevicegroup offline devicegroup-name scswitch -S -h node-name clnode evacuate node-name #1 Execute this command in step m in Set Up Solaris Cluster on page 3-51 of Preparations for a New Installation of HDLM on page 3-45 and in step 14 in Remove Solaris Cluster Settings on page 3-186 in Remove Solaris Cluster Settings on page 3-186.
A-4 Sun Cluster 3.
B Functional Differences Between Versions of HDLM This section explains the functional differences between newer and older versions of HDLM. □ Functional Differences Between Version 6.1 or Later and Versions Earlier Than 6.1 □ Functional Differences Between Version 6.0 or Later and Versions Earlier Than 6.0 □ Precautions on Differences in Functionality Between HDLM 5.6.1 or Earlier and HDLM 5.6.
Functional Differences Between Version 6.1 or Later and Versions Earlier Than 6.1 In a new installation of HDLM, the default load balancing algorithm has been changed from the Round Robin algorithm to the Extended Least I/Os algorithm. Functional Differences Between Version 6.0 or Later and Versions Earlier Than 6.0 • The format of the displayed HDLM version has been changed when HDLM commands or utilities are executed.
Acronyms and abbreviations The following acronyms and abbreviations might be used in this guide.
D DBMS Database Management System Dev Device DMP Dynamic Multipathing DNS Domain Name Server DR Dynamic Reconfiguration F FC Fibre Channel FC-SP Fibre Channel Security Protocol FO Failover FQDN Fully Qualified Domain Name G GAM Group Membership and Atomic Broadcast GMT Greenwich Mean Time GUI Graphical User Interface # A B C D E F G H I J K L M N O P Q R S Acronyms-2 Hitachi Dynamic Link Manager (for Solaris) User Guide T U V W X Y Z
H HBA Host Bus Adapter HDev Host Device HLU Host Logical Unit HTTP Hypertext Transfer Protocol I I/O Input/Output IP Internet Protocol iSCSI Internet Small Computer System Interface L LAN Local Area Network LDAP Lightweight Directory Access Protocol LDEV Logical Device LDoms Logical Domains LLT Low Latency Transport # A B C D E F G H I J K L M N O P Q R S T U V W X Y Z Acronyms-3 Hitachi Dynamic Link Manager (for Solaris) User Guide
LU Logical Unit LUN Logical Unit Number M MP Maintenance Pack N NAS Network Attached Storage NFS Network File System NTP Network Time Protocol O OBP OpenBoot PROM OS Operating System P P Port R RADIUS Remote Authentication Dial in User Service # A B C D E F G H I J K L M N O P Q R S Acronyms-4 Hitachi Dynamic Link Manager (for Solaris) User Guide T U V W X Y Z
S SAN Storage Area Network SCSI Small Computer System Interface SLPR Storage Logical Partition SMTP Simple Mail Transfer Protocol SNMP Simple Network Management Protocol SP Service Pack SSL Secure Sockets Layer SVP Service Processor U UFS UNIX File System V VTOC Volume Table Of Contents W WWN World Wide Name # A B C D E F G H I J K L M N O P Q R S T U V W X Y Z Acronyms-5 Hitachi Dynamic Link Manager (for Solaris) User Guide
Z ZFS Zettabyte File System # A B C D E F G H I J K L M N O P Q R S Acronyms-6 Hitachi Dynamic Link Manager (for Solaris) User Guide T U V W X Y Z
Glossary This glossary explains the terms used in this manual. A automatic failback A function for checking the status of paths at regular intervals, and automatically placing the status of a path recovered from an error into the Online status. If a path with an Offline(E) or Online(E) status recovers from an error, an automatic failback will place the path Online.
CLPR(Cache Logical Partition) A function supported by the Hitachi USP series, Universal Storage Platform V/VM series, Virtual Storage Platform series, VSP G1000 series, Hitachi AMS2000/AMS/WMS/SMS series, HUS100 series, and HUS VM for logically splitting up a cache. This function can split up a cache into parity groups in the storage system, so that other parity groups do not affect the cache performance.
F failback A function for placing the status of a path recovered from an error into the Online status, and then switching the access path to the path that was just recovered. failover A function for switching to another normal path if there is an error in the current access path, thereby enabling the system to continue to operate. FCode A program written by the interpreter language that is not dependent on the machine. The HBA that can connect to a boot disk might incorporate FCode for use.
host A generic term for both servers and clients. host device A logical area in a host LU. (See also: host LU, host device name) host device name A name assigned to a host device. A name of the logical device file for an HDLM device is assigned to a host device name. A host device name is assigned using the cUtXdY format.
- The Least Blocks algorithm - The Extended Least Blocks algorithm local boot disk environment An environment in which a boot disk is located on the host. logical device file Device files under /dev/dsk and /dev/rdsk. Note that the term logical device used here is not an LDEV. LU (Logical Unit) A logical unit that is a logical volume defined in the storage system, and with which the host can perform input or output operations. (See also: host LU) N node A server in a cluster.
owner path A path that passes through an owner controller for an LU in the Thunder 9500V Series, or Hitachi AMS/WMS series is in use. A path that passes through a controller other than an owner controller is called a non-owner path. When the Lightning 9900V series, Hitachi USP series, Universal Storage Platform V/VM series, Virtual Storage Platform series, VSP G1000 series, Hitachi AMS2000 series#, Hitachi SMS series#, HUS100 series#, or HUS VM is in use, all paths are owner paths.
an LU that has been reserved is given only to the host that issued the reservation, so the LU cannot be accessed from multiple paths (coming from multiple hosts) simultaneously. However, because of this, load balancing is not possible. S SAN (Storage Area Network) A high-speed network connecting hosts to storage systems. This network is independent of a LAN and is dedicated to data transfers.
# A B C D E F G H I J K L M N O P Q R S Glossary-8 Hitachi Dynamic Link Manager (for Solaris) User Guide T U V W X Y Z
Index Symbols removal 3-193 removing 3-133 upgrading Solaris 4-4 buffer size 3-150 changing the buffer size per monitoring interval duration 3-152 bus number 6-56 /etc/system file setup 3-49 /kernel/drv/sd.
offline 6-6 online 6-12 set 6-17 view 6-34 commands of HDLM (dlnkmgr) help 6-4 monitor 6-74 components HDLM system components 2-4 core logic component Critical 2-32 cUtXdYsZ 2-10 2-8 D default and recommended values 6-18 delete 6-80 deleting path to existing LU dynamic reconfiguration 4-47 Dev 2-3, 2-4, Glossary-2 Dev number Glossary-2 directory to which collected information is output 7-4 disk group creating 3-157 disk label 3-45 Displayed program information 6-36 displaying information 6-34 LU informati
functions of HDLM 2-1 G getras.tar.Z 7-6 Global Link Manager 2-41 glossary Glossary-1 H HAM environment Glossary-3 handling path error 5-3 HBA 2-4 HBA (Host Bus Adapter) Glossary-3 hbsa.tar.
installation 3-45 performing an upgrade installation or reinstallation 3-73 preparations for a new installation of HDLM 3-45 preparations for an upgrade installation or reinstallation of HDLM 3-73 installation-information settings file contents 7-34 INSTALLATION_SETTINGS 7-35 load balancing algorithms 2-14 load balancing is applied 2-12 local boot disk environment Glossary-5 migrating from boot disk environment 3-116 removing 3-191 log file 7-42 log filter 2-33 logical device file Glossary-5 logical device
O obtaining path information 5-5 program information 5-8 offline 2-16, 6-6 offline(C) status 2-20 offline(E) status 2-20 online 2-16, 6-12 online status 2-19 online(D) status 2-20 online(E) status 2-19 online(S) status 2-20 operating environment 6-17 setting 6-17 view [view operation] 6-41 operation name 8-3 operations 6-2 displaying format 6-4 of the HDLM command 6-2 OS 3-3 overview command 6-2 HDLM 1-1 owner controller 2-12, Glossary-5 owner path 2-12, Glossary-6 P P 2-4 path 2-3, Glossary-6 adding dynam
/kernel/drv/sd.
Hitachi Dynamic Link Manager (for Solaris) User Guide
Hitachi Data Systems Corporate Headquarters 2845 Lafayette Street Santa Clara, California 95050-2639 U.S.A. www.hds.com Regional Contact Information Americas +1 408 970 1000 info@hds.com Europe, Middle East, and Africa +44 (0)1753 618000 info.emea@hds.com Asia Pacific +852 3189 7900 hds.marketing.apac@hds.
