HP Serviceguard for Linux Version A.11.
Legal Notices The information in this document is subject to change without notice. Hewlett-Packard makes no warranty of any kind with regard to this manual, including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose. Hewlett-Packard shall not be held liable for errors contained herein or direct, indirect, special, incidental or consequential damages in connection with the furnishing, performance, or use of this material. Warranty.
Table of Contents 1 Serviceguard for Linux Version A.11.18 Release Notes...................................................................11 Announcements..................................................................................................................11 March 2008 Patches (A.11.18.03)...................................................................................11 Supported Platforms and Linux Distributions.............................................................
About Cross-Subnet Configurations.............................................................................30 Configuration Tasks.................................................................................................30 Restrictions...............................................................................................................30 For More Information..............................................................................................
For Red Hat:.............................................................................................................55 For SUSE:.................................................................................................................55 Installing the WBEM Provider 2.0.................................................................................56 Installing Quorum Server..............................................................................................
JAGag09818 (SR8606453031): Cannot modify SERVICE_HALT_TIMEOUT to default value (0 seconds)...........................................................................................................77 JAGag11398 (SR8606454772): cmcld aborts with Failed to set keep alive: Invalid argument.......................................................................................................................77 JAGag12644 (SR8606456223): cmsrvassistd doesn’t check script permissions.............
QXCR1000771266: Cluster Create: ASCII files are not distributed if cluster name contains $.......................................................................................................................85 Known Problems and Workarounds for Serviceguard Manager.................................85 JAGag35535 (SR8606481564): Adding a node to a cluster with Lock LUN fails.....86 JAGag33987 (SR8606479836): On Linux, SMH cluster summary status shows failed when a package is halted...........................
List of Figures 1-1 1-2 1-3 1-4 1-5 1-6 1-7 1-8 1-9 1-10 1-11 1-12 8 System Management Homepage (SMH) - Login........................................................42 System Management Homepage................................................................................43 System Management Homepage with Serviceguard Manager..................................44 Systems Insight Manager (SIM) - Login......................................................................46 Expanded Left Panel...............
List of Tables 1-1 1-2 1-3 1-4 1-5 1-6 Base Modules..............................................................................................................20 Optional Modules.......................................................................................................21 Legacy Configuration-File Parameters and Modular Equivalents.............................26 Legacy Control-Script Parameters and Modular Equivalents....................................27 Accessing Serviceguard Manager........
1 Serviceguard for Linux Version A.11.18 Release Notes Announcements This section announces the most important features and limitations of Serviceguard for Linux A.11.18. For more information, see “What’s in this Release” (page 15). March 2008 Patches (A.11.18.03) For information about new features introduced in the March 2008 patches, see “New Features for March 2008 Patches” (page 14). This patched version of Serviceguard for Linux is formally known as Serviceguard for Linux A.11.18.03.
NOTE: For full details of supported errata, etc., see the HP Serviceguard for Linux Certification Matrix at www.hp.com/info/sglx. Product Structure Serviceguard for Linux A.11.18 is available under the following product numbers: • T2391AA — A.11.18 — software and license for HP Integrity Servers — 1 license for HP Integrity server - one license required per processor — 1 CD — Installation and support services available through product options • B9903BA — A.11.
New Method for Configuring Packages Serviceguard A.11.18 introduces a new process for configuring packages. Packages created by this method are known as modular packages, because you can now build a package using only the sets of functions (modules) that this package needs; and you no longer need to create and distribute a separate package control script.
For the latest information on supported configurations, see the Serviceguard for Linux Certification Matrix at the address given under “What Documents are Available for This Version ” (page 48). Cluster Name Restrictions The following characters must not be used in the cluster name if you are using the Quorum Server: at-sign (@), equal-sign (=), or-sign (|), semicolon (;).
• • • • Red Hat 5 x86-64 SGLX_00206 SLES 10 IA32 SGLX_00207 SLES 10 IA64 SGLX_00208 SLES 10 x86-64 SGLX_00209 NOTE: These Serviceguard Manager patches include a fix for a defect, so you should install them if you use Serviceguard Manager, whether or not you intend to use the new Serviceguard features above. See “Patches for this Version” (page 76) for more information.
See the subsections that follow for details; see also the “Announcements” (page 11). For information about documentation, see “What Documents are Available for This Version ” (page 48). • There is a new, modular method of configuring packages; see “About Modular Packages” (page 18). — In addition, in the patches listed under “New Features for March 2008 Patches” (page 14), Serviceguard provides a means of automatically migrating existing (legacy) packages to modular packages.
executing this command outside the context of a package; see the entry for cmmodnet in Appendix A of Managing Serviceguard for Linux, and the cmmodnet manpage. — Serviceguard for Linux now supports IPv6 addresses, for data links only, and with some additional restrictions. See Appendix E of Managing Serviceguard for Linux for details.
• • A new command, cmdisklock, can be used to repair or replace a lock LUN; see “Replacing a Lock LUN” in Chapter 8 of Managing Serviceguard for Linux, and the cmdisklock manpage, for more information. Serviceguard for Linux version A.11.18 introduces initial support for two application interfaces based on the Service Availability Forum Specification (SAF, saforum.org). These are for use only by ISVs (Independent Software Vendors) with special support contracts.
IMPORTANT: Serviceguard does not force you to use this new methodology. Existing packages, known as legacy packages, continue to run without modification, and can be maintained by means of the older methodology; see “Configuring a Legacy Package” in Chapter 7 of Managing Serviceguard for Linux. You can also still create a new legacy package. If you are using a Serviceguard Toolkit such as Serviceguard NFS Toolkit, consult the documentation for that product.
Table 1-1 Base Modules 20 Module Name Parameters Comments failover package_name module_name module_version package_type node_name auto_run node_fail_fast_enabled run_script_timeout halt_script_timeout successor_halt_timeout script_log_file operation_sequence log_level failover_policy failback_policy priority Base module. Use as primary building block for failover packages.
Optional Package Modules Add optional modules to a base module if you need to configure the functions in question. Table 1-2 Optional Modules Module Name Parameters Comments dependency dependency_name dependency_condition dependency_location Add to a base module to create a package that depends on one or more other packages. monitor_subnet monitored_subnet monitored_subnet_access Add to a base module to configure subnet monitoring for the package.
Table 1-2 Optional Modules (continued) Module Name Parameters Comments external_pre external_pre_script Add to a base module to specify additional programs to be run before volume groups and disk groups are activated while the package is starting and after they are deactivated while the package is halting. external external_script Add to a base module to specify additional programs to be run during package start and stopped at package halt time.
NOTE: For more information, see the comments in the editable configuration file output by the cmmakepkg command, and the cmmakepkg manpage. See also Managing Serviceguard for Linux: “Package Configuration Planning” in Chapter 4, and Chapter 6 “Configuring Packages and Their Services” module_name The module name (for example, failover, service, etc.) Do not change it.
• • 4 - detailed debugging information 5 - function call flow priority Assigns a priority to a failover package whose failover_policy is configured_node. Valid values are 1 through 3000, or no_priority. The default is no_priority. Can be used in both modular and legacy packages.
The variable name must be in the form pev_ and contain only alphanumeric characters and underscores. The letters pev (upper-case or lower-case) followed by the underscore (_) are required. The variable name and value can each consist of a maximum of MAXPATHLEN characters (4096 on Linux systems). You can define more than one variable. For more information, see “About External Scripts” in Chapter 4 of Managing Serviceguard for Linux, and the comments in the package configuration file.
tables show, some appear under different names in modular packages, and in a few cases a legacy parameter has no modular equivalent (these are shown in bold). An asterisk in the table indicates that the formal requirements for the parameter value have changed (in terms of what characters are permitted, etc.). See “Parameters whose Formal Requirements Have Changed” (page 27).
Table 1-4 Legacy Control-Script Parameters and Modular Equivalents Legacy Parameter Modular Parameter SUBNET[] ip_subnet IP ip_address SERVICE_NAME[] service_name SERVICE_CMD[] service_cmd SERVICE_RESTART[] service_restart CONCURRENT_FSCK_OPERATIONS concurrent_fsck_operations CONCURRENT_MOUNT_AND_UMOUNT_ OPERATIONS concurrent_mount_and_umount_ operations FS_UMOUNT_COUNT fs_umount_retry_count FS_MOUNT_RETRY_COUNT fs_mount_retry_count LV[] fs_name FS[] fs_directory FS_TYPE[] fs_type
The value of each of these parameters must now be: any name, up to a maximum of 39 characters, that: • • • starts and ends with an alphanumeric character otherwise contains only alphanumeric characters or dot (.), dash (-), or underscore (_) is unique among package names in this cluster IMPORTANT: These rules apply to both legacy and modular packages. Restrictions on the values of these parameters in previous Serviceguard releases were less stringent.
About Multi-Node Packages Multi-node packages run simultaneously on more than one node in the cluster. Failures of package components such as applications, services, or subnets, will cause the package to be halted only on the node on which the failure occurred. Relocatable IP addresses cannot be assigned to multi_node packages. IMPORTANT: Multi-node packages must either use a clustered file system such as Red Hat GFS, or not use shared storage.
• • You can use cmmakepkg -l (letter “l”) to see a list of all available modules, including non-Serviceguard modules such as those supplied in the HP Toolkits. You can use cmgetconf -P write out the configuration information for all packages in the cluster to a specified location; for example: mkdir /tmp/pkgconfig cmgetconf -c clusterA -P /tmp/pkgconfig About Cross-Subnet Configurations As of Serviceguard A.11.
• Each heartbeat subnet on each node must be physically routed separately to the heartbeat subnet on another node; that is, each heartbeat path must be physically separate: — The heartbeats must be statically routed; static route entries must be configured on each node to route the heartbeats through different paths. — Failure of a single router must not affect both heartbeats at the same time.
IMPORTANT: Although cross-subnet topology can be implemented on a single site, it is most commonly used by extended-distance clusters. Design and configuration of such clusters are covered in the disaster-tolerant documentation delivered with Serviceguard. For more information, see the following documents at http://www.docs.hp.
• You cannot change the designation of an interface from STATIONARY_IP to HEARTBEAT_IP unless the subnet is common to all nodes. Remember that the HEARTBEAT_IP must be on the same subnet on all nodes, and must be an IPv4 address. • • • You cannot delete a primary interface without also deleting any standby interfaces, unless the standby is being used by another primary interface that is not being deleted.
NODE_NAME NETWORK_INTERFACE HEARTBEAT_IP #NETWORK_INTERFACE #STATIONARY_IP NETWORK_INTERFACE ftsys10 lan1 192.3.17.19 lan0 15.13.170.19 lan3 You can now uncomment the entries and apply the new configuration; see “Changing the Cluster Networking Configuration while the Cluster Is Running” in chapter 7 of Managing Serviceguard for Linux for more information.
NOTE: If you use the cmmigratepkg utility to convert a package that includes the SGLX_VM parameter, you will see a warning; for more information see “JAGag48881 (SR8606496963): cmmigrate does not handle virtual machine package script updates ” (page 79). • Serviceguard running on VMware virtual machines requires the sg_persist command, which you may need to install separately, depending on your platform and Linux distribution.
IMPORTANT: This represents a net addition to the time it takes for the cluster to re-form. For example, if the cluster typically took 40 seconds to re-form before any VM nodes were added, it will generally take about 80 seconds when one or more VM nodes are members of the cluster, if all those nodes have the hpvminfo or vminfo software. If any VM node without that software is a member of the cluster, it will take about 110 seconds.
Access changes as of A.11.16 Serviceguard version A.11.16 introduced a new access method. As of A.11.16, Serviceguard uses Access Control Policies, also known as Role-Based Access, rather than cmclnodelist or .rhosts, to authenticate users. For more information about Access Control Policies, see Chapter 5 of the the Managing Serviceguard for Linux manual (docs.hp.
About the Serviceguard Manager SMH Plug-In HP Serviceguard Manager B.01.01 is a web-based, HP System Management Homepage (HP SMH) plug-in application, that replaces the functionality of the earlier Serviceguard management tools. HP Serviceguard Manager allows you to monitor, administer and configure a Serviceguard A.11.18 cluster from any HP system with a web browser. For information on installing Serviceguard Manager, see “Installing and Setting Up Serviceguard Manager on Linux” (page 57).
Limitations • Using Serviceguard Manager, you can create only legacy packages. But you can monitor and administer both legacy and modular packages. See “About Modular Packages” (page 18) for more information about modular and legacy packages. • You cannot use Serviceguard Manager to configure cross-subnet clusters or packages (see “About Cross-Subnet Configurations” (page 30)). — The Serviceguard Manager Configuration pull-down menu is disabled in a cross-subnet configuration.
NOTE: For more information about DSAU, see the Distributed Systems Administration Utilities User’s Guide, version 1.4, at http://www.docs.hp.com -> Network and Systems Management Help Subsystem Once Serviceguard Manager is running, use the tooltips by moving your mouse over a field from the read-only property pages for a brief definition for each field. You can also access the online help by clicking the button located in the upper-right hand corner of the screen to view overview and procedure information.
Table 1-5 Accessing Serviceguard Manager Scenario Use Case Number of Clusters Serviceguard Version Serviceguard Manager Version For more information Scenario #1 1 A.11.18 B.01.01 (page 41) 1 or more cluster(s) A.11.16 A.05.01 (page 44) more than 1 cluster A.11.16 - A.11.18 On management server: (page 45) Single cluster management Scenario #2 Multi-cluster management Scenario #3 Web-based multi-cluster management A.05.01 with HP SIM 5.10 On each Serviceguard A.11.18 node: B.01.
Figure 1-1 System Management Homepage (SMH) - Login 2. When the System Management Homepage login screen appears, enter your login credentials and click Login. When the System Management Homepage for the selected server appears, click the name of the cluster link from within the Serviceguard Cluster box. 42 Serviceguard for Linux Version A.11.
Figure 1-2 System Management Homepage NOTE: If a cluster is not yet configured, then you will not see the Serviceguard Cluster section on this screen. To create a cluster, from the SMH Tools menu, you must click Serviceguard Manager link in the Serviceguard box first, then click Create Cluster. The figure below shows a browser session at the HP Serviceguard Manager Main Page.
Figure 1-3 System Management Homepage with Serviceguard Manager 1 2 3 4 5 6 7 Number Description 1 System Management Homepage (SMH) menus. 2 HP Serviceguard Manager pull-down menus. 3 HP Serviceguard Manager tabs. 4 Alert Summary. Click on the icon to see all cluster and package alerts. 5 Cluster status icons. 6 Use the check boxes to multi-select nodes or packages from which you can select an action to take from either the Administration or Configuration pull-down menus.
• • One or more clusters running Serviceguard version A.11.15 through A.11.16. Serviceguard Manager version A.05.01 installed on a Windows, HP-UX or Linux management workstation. Installation Requirements: Requires Serviceguard Manager installation on each management workstation. Launching Serviceguard Manager from Windows Workstations: There are 2 ways to launch Serviceguard Manager in Windows: • By default, a shortcut icon is placed on your desktop (labeled SG Manager). Click it to launch the program.
• • Serviceguard Manager B.01.00 and System Management Homepage (SMH) must be installed on each node of each cluster running Serviceguard version A.11.17.01 (for HP-UX only). Either the WBEM Provider or the SNMP sub-agent must be installed on each node of each cluster running Serviceguard for Linux version A.11.18. For installation instructions, see “Installing the SNMP Sub-Agent” (page 55) and “Installing the WBEM Provider 2.
Figure 1-5 Expanded Left Panel 4. Expand HP Serviceguard, and click on a Serviceguard cluster. The following screen appears if you click on a Serviceguard cluster running A.11.17.
Figure 1-6 HP Serviceguard Manager Homepage (SIM) What Documents are Available for This Version For information about the current version of Serviceguard, and about older versions, see the Serviceguard documents posted at http://docs.hp.com/ -> High Availability -> Serviceguard for Linux. Documents for related products are also on the High Availability page. The following documents relate to Serviceguard A.11.18 and related high availability products.
• • • • Serviceguard, SGeRAC, and SMS Compatibility and Feature Matrix, on docs.hp.com -> High Availability, under the heading Serviceguard -> Support Matrixes. Serviceguard Manager Support Matrix on docs.hp.com -> High Availability, under the heading Serviceguard -> Support Matrixes. Serviceguard for Linux Order and Configuration Guide at www.hp.com/info/ sglx. This contains the latest information about supported configurations. HP Serviceguard for Linux Certification Matrix at www.hp.com/info/sglx.
NOTE: All nodes must be on the same distribution, including errata level and architecture version (i.e. ia32, i686, x86_64, ia64). Using different distribution versions within the same cluster is supported only during rolling upgrade. Servers using processors from Intel and AMD can be mixed in a cluster. For supported rolling upgrade paths see “Rolling Software Upgrades”.
• • • • • • • hacl-gs 5301/TCP HA Cluster General Services hacl-cfg 5302/TCP HA Cluster TCP configuration hacl-cfg 5302/UDP HA Cluster UDP configuration hacl-probe 5303/TCP HA Cluster TCP probe hacl-probe 5303/UDP HA Cluster UDP probe hacl-local 5304/TCP HA Cluster commands hacl-test 5305/TCP HA Cluster test The ports reserved for authentication are also used by Serviceguard: • • auth 113/TCP authentication auth 113/UDP authentication In addition, Serviceguard also uses dynamic ports (typically in the r
All nodes in the cluster must allow the following communications: • from the remote nodes: — TCP on ports 5302 - and allow only packets with the SYN flag — UDP on port 5302 • to the remote nodes: — TCP and UDP on port numbers 49152-65535 The remote nodes must allow the following communications: • from the cluster nodes — TCP and UDP on port numbers 49152-65535 • to the cluster nodes — TCP on ports 5302 - and allow only packets with the SYN flag — UDP on port 5302 Authentication communication must allo
Installing HP Serviceguard for Linux IMPORTANT: Before you begin installing Serviceguard, make sure that all components of the cluster have been upgraded to their latest firmware versions. NOTE: Serviceguard for Linux depends on the xinetd service. Make sure that the xinetd rpm has been installed from the distribution media and is enabled.
2. For versions earlier than Red Hat 5: up2date --configure Add authd* to the list of items to exclude (removeSkipList in the text-based version of up2date, or theExclude tab in the GUI). This will prevent up2date from trying to install new versions of authd in future. For Red Hat 5 and later versions: Edit /etc/yum.conf and add authd to the exclude list. For example, in the [main] section, add: exclude=authd Installation 1.
SUSE Installation 1. On the Serviceguard for Linux CD, change to the appropriate directory: cd /Serviceguard/ For example cd SLES10/Serviceguard/IA64 2. Install the sgcmom and serviceguard rpms; for example, on an HP Integrity system: rpm -ihv sgcmom-B.05.00.00-0.sles10.ia64.rpm rpm -ihv serviceguard-A.11.18.00-0.sles10.ia64.rpm 3.
Installing the WBEM Provider 2.0 Proceed as follows if you need to install the WBEM Provider. (For example, you may need the SNMP sub-agent for use with Serviceguard Manager and HP SIM; see “Scenario 3 - Web-based multi-cluster management” (page 45).) For Red Hat: 1. Make sure you have the Pegasus software on your system: rpm - qa | fgrep tog-pegasus If Pegasus is not on the system, install it from your Red Hat installation source (for example, your installation CDs). 2.
Installing Quorum Server If you choose to use a Quorum Server rather a lock LUN for tie-breaking, install the Quorum Server software on a system outside the cluster. The Quorum Server software is on the Serviceguard for Linux CD. Updated versions may be available from http:// www.hp.com/go/softwaredepot ->High Availability ->ServiceGuard Quorum Server. See the Quorum Server Version A.03.
NOTE: For Red Hat 5 you need to install additional software from your distribution source (for example the Red Hat 5 installation CDs): • For Red Hat 5 (i386) locate and install libXp-1.0.0-8.i386.rpm, for example: rpm -Uhv libXp-1.0.0-8.i386.rpm • For Red Hat 5 (IA64) locate and install libXp-1.0.0-8.ia64.rpm, for example rpm -Uhv libXp-1.0.0-8.ia64.rpm • For Red Hat 5 (x86-64) locate and install libXp-1.0.0-8.i386.rpm, for example rpm -Uhv libXp-1.0.0-8.i386.
NOTE: To obtain DSAU for Red Hat 5, you must download it from http:// www.hp.com/go/softwaredepot -> High availability -> Linux Distributed System Administration Utilities. Select Receive for free and choose the version for your platform and Linux version. Downloading and Installing the Required Software from the Internet • For Integrity platforms: — Download the SMH and Tomcat software from http://www.hp.com/go/softwaredepot -> Linux.
Installing HP SMH and HP SMH Tomcat • For Integrity platforms, proceed as follows: 1. Navigate to the location where you downloaded the SMH software (see “Downloading and Installing the Required Software from the Internet” (page 59)); for example: cd /tmp 2. Use the tar command to extract the file as follows (all one line): tar -xvf SupportPack--..tar For example: tar -xvf SupportPack-4.50-sles10.ia64.tar 3.
or rpm -Uhv hpsmh-.linux-release..rpm For example: rpm -Uhv hpsmh-2.1.7-168.linux.i386.rpm NOTE: If you are installing the SUSE version, you can ignore the message that may appear warning you to run SMHCONFIG. 3. Install the Apache Tomcat software, as follows (all on one line): rpm -Uhv hpsmh-.linux..rpm For example: rpm -Uhv hpsmh-tomcat-1.0-11.sles10.rpm • 1.
rpm -Uhv sgmgrpi-B.01.01-3.sles10.IA32.rpm 3. Restart SMH to integrate SMH with Serviceguard Manager: /etc/init.d/hpsmhd restart NOTE: See “Launching Serviceguard Manager” (page 40) for more information about launching Serviceguard Manager. Installing and Configuring the Distributed Systems Administration Utilities (DSAU) This section explains how to get, install and configure consolidated logs for a cluster, and set up a shared storage location.
1. Choose from the list of disks to be configured for shared storage: fdisk -l The following output appears. Choose a disk accessible from all the members of the cluster. It should be of type 8e Linux LVM. For example, in this case it should be either /dev/sda1 or /dev/sdb1. Disk /dev/cciss/c0d0: 72.
NOTE: On HP Integrity servers, sfdisk does not exist. Run vgscan, then if vgscan does not find the volume group, you will need to reboot the node. 10. Back up the configuration: vgcfgbackup /dev/vgclog Using the Command Line to Configure the Package clog for Consolidated Logging When configuring the package clog in a cluster, you can either create it as a server, which receives logging from all the configured clients, or as a client, which forwards syslog files and package logs to the server.
Enter the TCP port to be used for receiving logs []: 1009 Would you like to consolidate this cluster's syslogs? (y/n) [y]: y Would you like to consolidate this cluster's package logs? (y/n) [y]: y 3. When you have answered the questions, the package will be created and started on the primary node. Using Serviceguard Manager to Configure the Package clog for Consolidated Logging 1. 2. 3.
Rules 1. 2. 3. Serviceguard for Linux must be halted (not running) on the node from which the rpm -e command is issued. The node from which the rpm -e command is issued must not be a member of any cluster. The rpm -e command should be issued from one system at a time. That is, if Serviceguard for Linux is being uninstalled from more than one system, it should be removed from one system at a time.
• • Make sure that all applications that run on the current OS will also be supported with the new OS. Verify that the new OS supports the current cluster hardware configuration and drivers (network interfaces, bonding driver, shared disk storage. etc.). NOTE: If there is no HP bonding driver for the Linux version you are using, use the bonding driver supplied with the Linux distribution. • Do a full backup on each node.
• • • • All nodes must be running the same releases of Linux and Serviceguard before the upgrade. Rolling upgrades are not intended as a means of using mixed releases of Serviceguard or Linux within the cluster. It is highly recommended that you upgrade all cluster nodes as quickly as possible to the new release level. You cannot delete Serviceguard software (via rpm -e) from a node while the cluster is in the process of rolling upgrade.
NOTE: Make sure that all your network and storage interfaces are supported by the new OS. 3. Record the storage configuration, such as all LVM information, and if possible, collect a list of hardware disks configured, for example sfdisk -l On SUSE Linux Enterprise Server, you may need to run YAST or YAST2. 4. Back up the following files on media that can be easily recovered by the node after its upgrade or a new OS installation: • Host files: /root/.
5. Verify that all disks and their filesystems are the same prior to this OS upgrade or new installation. Check and compare with the disk layout collected before the upgrade. Use command vgscan to make sure the node with new OS sees all its previously configured LVM disks. 6. Install the new Serviceguard release. Follow the instructions under “Installing HP Serviceguard for Linux” (page 53). The following examples show only the Serviceguard rpm.
The deadman driver sources and associated files are located at $SGROOT/drivers. The pidentddriver is located at: /usr/src/pidentd-/driver where version represents the version number for the pidentd that was last installed. Follow the instructions in the README file in the directory of each driver. If you have installed a new OS version, you must run a convert program. This will convert the binary file (cmclconfig) to its new release format.
NOTE: This and the following figures are from a use case in which the starting point of the upgrade was Serviceguard A.11.14.02 and Red Hat AS 2.1, and the rolling upgrade was to Serviceguard A.11.15.02 and Red Hat EL3. This is for illustration only; the upgrade process has not changed since. Figure 1-7 Running Cluster Before Rolling Upgrade Node1 Pkg1 SG A 11.14.02 Red Hat AS 2.1 Node2 Pkg2 SG A.11.14.02 Red Hat AS 2.1 Step 1. Record all data as specified in the preparation section.
Figure 1-8 Running Cluster with Packages Moved to Node 2 Node1 Node2 SG SG A.11.14.02 Pkg1 A.11.14.02 AS 2.1 Pkg2 AS 2.1 Step 2. Upgrade node1 to the new operating system release (in this example, it was Red Hat EL 3), and install the new version of Serviceguard, as shown in Figure 1-9. Figure 1-9 Node 1 Upgraded to Red Hat EL 3 Node1 Node2 Pkg1 Red Hat EL 3 Pkg2 SG A.11.14.02 Red Hat AS 2.1 Step 3.
Figure 1-10 Node 1 Rejoining the Cluster Step 4. Repeat the process on node2. Halt the node as follows: # cmhaltnode -f node2 This causes both packages to move to node1. Then upgrade node2 to the new Linux distribution and the new version of Serviceguard. Figure 1-11 Running Cluster with Packages Moved to Node 1 Node1 Pkg1 Pkg2 Node2 SG SG A.11.15.02 A.11.15.02 Red Hat EL 3 Red Hat EL 3 Step 5. Move pkg2 back to its original node.
# cmrunpkg -n node2 pkg2 # cmmodpkg -e pkg2 The cmmodpkg command re-enables switching of the package, which is disabled by the cmhaltpkg command. The final running cluster is shown in Figure 1-6. Figure 1-12 Running Cluster After Upgrades Node1 Pkg1 Node2 SG Pkg2 A.11.15.02 Red Hat EL 3 SG A.11.15.
Preventing Boot-Time vgscan and Ensuring Serviceguard Volume Groups Are Deactivated By default, Linux will perform LVM startup actions whenever the system is rebooted. These include a vgscan (on some Linux distributions) and volume group activation. This can cause problems for volumes used in a Serviceguard environment (for example, a volume group for a Serviceguard package that is not currently running may be activated). To prevent such problems, proceed as follows on the various Linux versions.
IMPORTANT: If you use Serviceguard Manager, you should install the Serviceguard Manager patches listed in that section, whether or not you intend to use the new features, as they fix a defect in the way Serviceguard Manager handles package-dependency configuration. Before installing Serviceguard, you should also check the Hewlett-Packard IT Resource Center web page for any new patch requirements: http://itrc.hp.com (Americas and Asia Pacific)http://europe.itrc.hp.
Failed to set keep alive: Invalid argument What is the solution? The code has been modified to close the TCP connection upon failure to set keep-alive with EINVAL, instead of aborting cmcld. This should result in re-establishing the connection; in the worst case a cluster re-formation could result.
The real problem is that IPv6 is not enabled: cmmodnet did not check whether IPv6 was enabled on the system before trying to add the address. • What is the solution? cmmodnet now checks whether IPv6 is enabled on the system before trying to add the address. JAGag48606 (SR8606496688): cmmigratepkg missing PEV and external script permission • What was the problem? The-e option of cmmigratepkg should generate Package Environment Variables (PEVs) but it does not.
Warning: at line: 541 function persist_reservation is not a Serviceguard function. Warning: at line: 629 function persist_release is not a Serviceguard function. • What is the solution? cmmigratepkg now handles the SGLX_VM parameter correctly. QXCR1000581069: Deleting multiple packages failed if a selected package was not down • • What was the problem? Deleting multiple packages failed if a selected package was not down.
• the Tools tab, because there are incompatibilities between HP SMH components and Serviceguard Manager B.01.01. What is the solution? Serviceguard Manager’s CGI script has been updated to be compatible with newer versions of Linux IA64 SMH. This fix requires that the following software is installed before installing the Serviceguard Manager patch: jrockit-R27.4.0-jdk1.5.0_12-linux-ipf.bin; the latest version of HP SMH (version 2.1.10-04 or later); and the latest version of HP SMH Tomcat (1.0-20 or later).
NOTE: If you need to build a modular package with VMware support, you must upgrade all nodes in the cluster to 11.18.03. Then you can use cmmakepkg and cmapplyconf to create the package. Known Problems and Workarounds This section lists known problems in Serviceguard Version A.11.18 at the time of publication. This list is subject to change without notice. More-recent information may be available on the Hewlett-Packard IT Resource Center: http://www.itrc.hp.com (Americas and Asia Pacific) or http://www.
6. 7. • • Re-apply the configuration Re-start the package Another solution is to add a new package with a service name matching that originally deleted and then halt the service with cmhaltserv. This would allow the problem to be resolved without halting either the package, node or cluster. Alternatively, if the cluster/node were halted and re-started the problem would go away. This resolution requires cluster/node downtime, however, and that is not always possible in 24x7 environments.
program cmclconfd is using a deprecated SCSI ioctl, please convert it to SG_IO cmclconfd uses SCSI ioctl calls to the lock LUN device to get various pieces of information. The ioctl currently being used is being deprecated but still works correctly. What is the workaround? (This is an informational message only and can safely be ignored.
QXCR1000766057: cmviewcl package state “blocked” is not documented • • What is the problem? cmviewcl shows the per-node state of multi-node packages as blocked, but the manpage does not explain what “blocked” means. What is the workaround? “Blocked” means that the multi-node package has never run on the node. This may be because a dependency has not been met, or auto_run is set to no.
JAGag35535 (SR8606481564): Adding a node to a cluster with Lock LUN fails • • What is the problem? Adding node to an existing cluster that has Lock LUN configured fails. What is the workaround or available patch? Use the command-line command cmquerycl -L for each new node.
JAGag05524 (SR8606448264): When editing roles and monitored resources, some buttons may be disabled • What is the problem? When configuring a cluster or a package: 1. 2. 3. • A user may forget to press the Add button after entering values in the form. The data entered in the form is lost. If an incorrect value is entered in the form, no error message is displayed when [Add] is pressed. The error message is displayed only when a different tab is selected.
JAGag35014 (SR8606480970): On Linux, only one Consolidated Log can be viewed at a time • • What is the problem? If you open 2 log windows on Linux, you will get 2 log windows with identical contents. You cannot view both the system log and the package log, on different windows, at same time. Only one log can be viewed at one time. What is the workaround or available patch? Open and view one Consolidated Log at a time.
