HP NetRAID-4M Configuration and Upgrade Guide (Release 5) Online Version March 2003
Notice The information contained in this document is subject to change without notice. Hewlett-Packard makes no warranty of any kind with regard to this material, including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose. Hewlett-Packard shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material.
Contents Chapter 1 Documentation Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Overview of This Manual . . . . . . . . . . Installing a NetRAID-4M in a Cluster . . . Operating System Support. . . . . . . . . Matched Set of Software Components . . Cable Installation . . . . . . . . . . . . . Release 4 Issues and Problem Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Contents Chapter 5 HP Server BIOS and Slotting Rules . . . . . . . . . . . . . . . . . . . . . . 17 BIOS and Slotting Requirements . . . . . . . . . . . . . . . . . . . . 17 Comments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Chapter 6 Disk Drive Firmware Revision . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Firmware for HP Drives. . . . . . . . . . . . . . . . . . . How to Upgrade Drive Firmware . . . . . . . . . . . . . . To Upgrade Drive Firmware: . . . . . . . . . . .
1 Documentation Overview Installation Guide Supplement Installation Guide • Battery subsystem • BIOS utilities • Clusters • Controller software • Device drivers • Firmware flashing • Hardware • Overview • Software FAST User’s Guide • Command prompt window • Container • • • • • • • • • • • • • reconfiguration Controllers Disks Disksets & resources Enclosures Features/benefits GUI Interface NetWare via windows client Performance Remote management Storage concepts/ terminology Storage in a cluster Windows
Chapter 1 Documentation Overview Overview of This Manual Installing a NetRAID-4M in a Cluster Jumper settings may need to be changed before you install the NetRAID-4M card into a cluster configuration. See Chapter 2‚ SCSI Devices and Clustering. Operating System Support The HP NetRAID-4M is only supported with a specific set of operating systems. Some of those operating systems have special requirements and information that should be reviewed. See Chapter 3‚ Operating System Support for details.
2 SCSI Devices and Clustering The design of the HP NetRAID-4M requires the manual override of automatic SCSI termination for the cluster environment. This guarantees that the controller will still be able to terminate the SCSI bus even if the controller is powered off. Manual override requires: • removing the card, • possibly inserting a jumper, and • reinstalling the card.
Chapter 2 SCSI Devices and Clustering Channel 0 terminator jumper (J2) Channel 2 terminator jumper (J7) Channel 1 terminator jumper (J1) Channel 3 terminator jumper (J8) Figure 2-1.
3 Operating System Support HP NetRAID-4M supports the following operating systems: • Windows NT 4.0 with SP6a • Windows NT 4.0 Clustering • Windows 2000 • Windows 2000 Clustering • Linux (Please see HP web site for specific distributions and versions) • NetWare 5.1 with SP2 • NetWare 6 • Caldera UnixWare 7.1.1 • SCO OpenServer 5.0.6 NOTE HP Servers running SCO OpenServer can support no more than 4 HP NetRAID-4M controllers.
Chapter 3 • Operating System Support The Caldera updated version of UnixWare 7.1.1 and later, which incorporates all current patches to date. The Caldera part number for the current version is CC480-UW79-7.1.1. 1. Previous versions of UnixWare 7.1.1 with Maintenance Set 711011 installed. NOTE 6 As of this writing current patches for Caldera UnixWare can be downloaded at: http://www.caldera.com/support.
4 NetRAID-4M Software NetRAID-4M release 5 software components are available on the HP web site. These components (firmware, drivers, and utilities) are designed and tested to work together. Mismatched sets are not supported for the following reasons: • Each set of components is created in a build process that verifies the integrity of the files and the completeness of the components. • Each build is tested with an integrated set.
Chapter 4 NetRAID-4M Software Using the HP NetRAID-4M FAST utility, do the following: 1. Select Controller view from the View menu. 2. Select the controller. 3. Select Properties from the Edit menu. The Controller Properties dialog box will appear. The left-most column of Table 4-1 shows the software items that may need to be updated. The right-most column shows the correct build number for that component. Table 4-1.
Chapter 4 NetRAID-4M Software 2. Once the controller is opened, issue this command: controller details This command will display the build numbers of the components you have installed under the Component Revisions header. Compare the build number information displayed on the screen to that in the right-most column of Table 4-2 to determine if an upgrade is needed. Table 4-2.
Chapter 4 NetRAID-4M Software Table 4-2. Build Numbers by Release Component Name as displayed in CLI Release 5 Build # Software Component to Upgrade CLI API 4628 Utility Miniport driver N/A Driver (Linux drivers can be obtained from your Linux vendor.
Chapter 4 NetRAID-4M Software 2. Select HP NetRAID-4M Controller. 3. Select Download Drivers & Software. Update the NetRAID-4M Software The update process consists of 7 steps, as numbered below and explained in detail in the subtopics following. This order should be followed to ensure a successful update and to minimize rebooting. 1. Back up the system. 2. Remove the HP NetRAID-4M cluster component (if in a cluster environment). 3. Remove the HP NetRAID-4M FAST and/or CLI utilities. 4.
Chapter 4 NetRAID-4M Software Step 3-Remove the HP NetRAID-4M FAST and/or CLI utilities. NOTE During the operation of removing or installing the HP NetRAID-4M Utility from Windows NT and Windows 2000 both the FAST utility and the CLI utility will be removed or installed. Windows NT and Windows 2000 1. At the main Windows desktop, press Start > Settings > Control Panel. 2. Select Add / Remove Programs. 3. Select NetRAID-4M and press the Remove/Change button to delete the old FAST Utility program. 4.
Chapter 4 NetRAID-4M Software 7. The Update Device Driver Wizard will appear. Click Next. NOTE Do not let the installation utility search for the driver. This will result in it finding an existing out-of-date driver on the hard disk drive, rather than the newly downloaded driver. 8. Select the radio button: Display a list of the known drivers for this device so that I can choose a specific driver. Click Next. 9. Select Have Disk and Browse to the location where the driver was downloaded. 10.
Chapter 4 NetRAID-4M Software NetWare, Linux, UnixWare, OpenServer For instructions on updating the HP NetRAID-4M driver for operating systems other than Windows NT and Windows 2000, please see that operating system’s documentation. Step 5-Update the HP NetRAID-4M firmware 1. Insert 4M controller firmware update diskette 1 and reboot the server. 2. Choose Update Flash Image on the menu. 3. Choose to update from a:\ and choose OK. 4. The update requests diskette 2.
Chapter 4 NetRAID-4M Software NetWare 1. From the previously created diskette, copy the nwappdl.exe file to a Novell NetWare client system. 2. Run the nwappd1.exe program from the NetWare client. This uncompresses the files. 3. Copy all the uncompressed files from the NetWare client to the NetWare server. 4. Record the name of the directory that you copied the files to for future reference. The sys:system\ directory is recommended. 5.
Chapter 4 16 NetRAID-4M Software
5 HP Server BIOS and Slotting Rules BIOS and Slotting Requirements Table 5-1 summarizes the operating requirements for the successful installation of the HP NetRAID-4M in the supported HP Servers for Release 5. For further information regarding the slotting preferences for each HP Server, please refer to “Comments” on page 18. Table 5-1.
Chapter 5 HP Server BIOS and Slotting Rules Table 5-1. Supported HP Server Environments for the HP NetRAID-4M, Release 5 Slotting Preferences Key: - Excluded Slot Server Comments x Preferred Slot OK Slot Minimum BIOS LT 6000 8 4.06.43 PW 2 LT 6000 u3 8 4.06.43 RL 2 LXr 8000 9 Prod 12 4 LXr 8500 10 Prod 10 tc 4100 11 4.06.
Chapter 5 HP Server BIOS and Slotting Rules 9. Use Slots 7-10 because they are 64 bit slots. Slots 1-6 are 32 bit slots, and 1-2 are typically used for the NIC and RMC. 10. Use Slots 3-10 because 1-2 are typically used for the NIC and RMC. NOTE The maximum number of HP NetRAID-4M cards is four while using NetWare 5.X and 6X. 11. Use Slots 1-2 because they are 64 bit slots. Slots 1-6 are 32 bit slots. 12. The Remote Management Card for this server only works in this slot.
Chapter 5 20 HP Server BIOS and Slotting Rules
6 Disk Drive Firmware Revision Firmware for HP Drives It is important that HP disk drives have the newest available firmware to help insure mass storage subsystem stability. The newest available hard disk drive firmware and update utilities are available at: http://h20004.www2.hp.com/soar_rnotes/bsdmatrix/matrix65146en_US.html First check if any of the hard disk drives listed below are installed.
Chapter 6 WARNING Disk Drive Firmware Revision There should be no other disk activity when making a firmware upgrade. Make sure that the activity lights on the drives indicate no I/O activity before you do a firmware update on the drives. Method 1 - Flexible Diskette Update 1. Download the HP SCSI Hard Disk Drive Firmware Update Tool package (Uthdd.exe) 2.
Chapter 6 Disk Drive Firmware Revision Special Firmware Update Required for Certain IBM Drives The SCSI Detail Agent has exposed a hard disk drive firmware issue. Errors occur if you are using any Ultra3 hard disk drives listed in Table 6-1, and you are using firmware D84D or P84D along with Netserver Agents 4.53 or 4.53a. Table 6-1.
Chapter 6 Disk Drive Firmware Revision Verifying the Firmware Revision: From Windows Start, choose Settings, then Control Panel, and then SCSI Adapters. Double click on the appropriate adapter. Drive information will appear and the last portion is the firmware revision. For example, HP 9.10GB C 80-D84D. Workaround The workaround for all of these issues is to disable the SCSI Detail Agent. Disabling this agent will not require any reboots.
Chapter 6 Disk Drive Firmware Revision Solution The SCSI Detail Agent (included in the Netserver Agents 5.00 or later, which are currently available on the web) has been modified to not send Log Select or Log Sense commands to disk drives with D84D or P84D firmware. This modification to the SCSI Detail Agent is included on the L.19.00 HP Netserver Navigator CD. Also, the new disk drive firmware D94N or P94N will resolve this issue for disk drives with D84D or P84D firmware.
Chapter 6 26 Disk Drive Firmware Revision
7 Rack Storage/12 Revision Rack Storage/12 Revision Level Required Your RS/12 revision level must be correct and current. Table 7-1. Rack Storage/12 Revision History RS/12 Accessory Replacement PCA Rev SAF-TE Status D5989A D6025A D6025-63001 5064- B0 6617 1.01 Obsolete D5989B D6025B D6025-63002 5064- C0 7944 1.02 First release D5989B D6025B D6025-63003 5064- D 7944 1.02 Resistor fix D5989C D6025C D6025-63004 5065- C0 2726 1.03 First release D5989C D6025C D6025-63005 5065- C0 2726 1.
Chapter 7 Rack Storage/12 Revision Make an RS/12 Firmware Update Diskette Run the rs12fw.exe program and follow the instructions to create a RS/12 firmware update diskette (you will need a blank, formatted floppy disk). Update the Firmware 1. Be sure you have a current system backup before proceeding. 2. Shut down your operating system. 3. Boot to the RS/12 Firmware Update floppy diskette. 4.
8 Internal Cabling HP Netserver LC 2000: Installing the New SCSI Cable Connecting Common Tray Devices Connect the new SCSI cable (5185-2359) to the devices in the Common Tray as shown in Figure 8-1: Figure 8-1. HP Netserver LC 2000 Common Tray Cabling* *The slot location shown in Figure 8-1 is for illustration only. Other slot locations will work. (For maximum performance, install your HP NetRAID-4M in a 64-bit slot.) 1. Remove the Netserver from the rack enclosure.
Chapter 8 Internal Cabling 2. Remove the system cover. (Refer to the HP Netserver LC 2000/2000r Installation Guide.) 3. Install the HP NetRAID-4M in a PCI slot. (Refer to the documentation supplied with the HP NetRAID-4M for more detailed instructions.) 4. Install the new SCSI cable. 5. Connect the non-terminated end of the new SCSI cable to the internal LVD/SE connector on the HP NetRAID-4M. Push the connector until it clicks into place. 6.
Chapter 8 Internal Cabling HP Netserver LH 3000/LH 6000: Installing the New SCSI Cable Connecting Common Tray Devices Connect the new SCSI cable (5185-2359) to the devices in the Common Tray as shown in Figure 8-2: Clip Clip Floppy Drive Terminator HBA CD-Rom PSI Slot Bay 1 PSI Slot Bay 2 PSI Slot PSI Slot PSI Slot PSI Slot PSI Slot Power Supply Figure 8-2. HP Netserver LH 3000/LH 6000 Common Tray Cabling* *The slot location shown in Figure 8-2 is for illustration only.
Chapter 8 Internal Cabling 6. Connect the terminated end of the cable that has the terminator block with the devices located in the Common Tray. (Start with the connector at the terminated end of the cable.) 7. Ensure that all cables are out of the way, and then replace the system cover. 8. After all covers have been replaced and secured, reconnect the power cord and all external cables as described in your documentation.
Chapter 8 Internal Cabling 3. Insert the new cable through the opening of the mid-plane divider and pull it to the other side. 4. Connect one end of the new SCSI cable to the internal LVD/SE connector on the HP NetRAID-4M. Push the connector until it clicks into place. 5. Connect the other end of the cable to the 68-pin connector on the right-side hotswap cage. Push the connector until it clicks into place. 6. Ensure that all cables are out of the way, and then replace the system cover. 7.
Chapter 8 34 Internal Cabling
9 Issues and Problem Resolution The issues and problem resolutions identified and resolved that affected all operating system platforms are listed below by operation environment, and then explained by issue. To find additional information about NetRAID-4M see HP’s web site at: http://netserver.hp.
Chapter 9 Issues and Problem Resolution 17: SMART MRIE Level 2 F F F F F F 18: Drive firmware download process can cause drives to go offline R R R R R R 19: Rack Storage/12 and voltage regulation R R R R R R 20: Rack Storage/12 with internal twisted pair cables R R R R R R 21: Rack Storage/12 new SAF-TE firmware to fix false flashing red status LED R R R R R R 22: Boot order considerations W W W W W W 23: 32 KB option ROM function W W W W W W Linux R R R R R R UnixWare 16: Cannot add 9 G
Chapter 9 Issues and Problem Resolution Linux W UnixWare 35: Linux open source driver needs patch to fix memory leak OpenServer Mapping of Issues to Operating Environments: F = new feature R = resolved W = workaround = not applicable Windows NT4 Windows 2000 ISSUES Novell NetWare Table 9-1.
Chapter 9 Issues and Problem Resolution 55: Free space and file system extension W W Linux W UnixWare 54: Upgrading Windows 2000 NetRAID-4M driver OpenServer Mapping of Issues to Operating Environments: F = new feature R = resolved W = workaround = not applicable Windows NT4 Windows 2000 ISSUES Novell NetWare Table 9-1.
Chapter 9 Issues and Problem Resolution Problem: The Automatic Failover option gets executed regardless if the drive has been replaced or not. When a drive reports a failure, the rebuild process should happen onto a spare drive. If there is no spare drive assigned and the Automatic Failover option is enabled, the rebuild should happen onto the same SCSI ID, providing that the drive has been replaced. The UNCONFIG bit gets set when a new drive is reinserted.
Chapter 9 Issues and Problem Resolution Problem: The test case involves running both a “proprietary test” and “winthrax” to separate containers. After running in the time frame of 30min to several hours, the “proprietary test” detects incorrect data. An analysis of the data shows that the preceding 1 k write to the given chunk wrote the first 512 B correctly but the second 512 B was not written at all; that is, the second half of the 1 KB chunk had data that was written from the 2nd preceding write.
Chapter 9 Issues and Problem Resolution Mfg diagnostic scsi_x.exe encounters the problem by trying to do a bus scan on all possible host adapter numbers. When it does it at a nonexistent HA# after the inquire command is executed it enters an infinite loop, polling and waiting to see a command done status, but this never occurs. The driver should treat the inquire command to non-existing HA# as if no device found and returned the same command done status.
Chapter 9 Issues and Problem Resolution Problem: The PASSTHRU command s/g 3-4GB mapping routine added in an earlier build was not checking for a ZERO s/g count and was thus processing a nonexistent s/g list resulting in a infinite loop in the kernel code. Impact to customer: Serious. Controller can hang causing data unavailability. Resolution: Pass-thru code was fixed to check for zero SG count.
Chapter 9 Issues and Problem Resolution Resolution: By disabling the cache after the first drive failure, a second drive failure will no longer hang the system. 14 1 TB expansion Problem: A container expansion to 1 TB can cause a controller panic. Impact to customer: Serious. The user will be unable to expand the current container without having to rebuild the container and restoring the data from a current backup.
Chapter 9 Issues and Problem Resolution Problem: One of the new features that disk drives offer is SMART (Self-Monitoring And Reporting Technology). The new SMART compliant disk drives support informational exception conditions, including the SCSI sense codes FAILURE PREDITION THRESHOLD EXCEEDED, or WARNING. There are 7 MRIE (Method of Reporting Informational Exceptions): 0 - The device will not report exception conditions. 1 - (Not Supported.) 2 - Report exception before a SCSI command.
Chapter 9 Issues and Problem Resolution Example: On some servers (such as the LH 6000) the hard disk drive’s red LED will blink; on others (such as the LPr), the red LED will not be seen at all. Resolution: Disable background consistency check prior to HDD firmware upgrade. 19 Rack Storage/12 and voltage regulation Problem: There is a data-rate sensitivity in some Adapter Boards used in RS/12 enclosures. The symptom is that random drives in the RS/12 go offline at random times.
Chapter 9 Issues and Problem Resolution Problem: Customers may experience intermittent occurrences of hard disk drives going offline or not being recognized by the disk controller, system hangs, or SCSI bus parity errors on their HP D5989A/B/C Rack Storage/12. In some situations, the HP Rack Storage/12 systems may be missing two internal twisted pair cables.
Chapter 9 Issues and Problem Resolution Problem: An erroneous entry for a card (a “phantom” card) may be listed in the “Hard Drive Priority” screen in the server's BIOS. Other symptoms: the system will not boot from the floppy drive or the system hangs. Workaround: You may need to use the NVRAM / Clear CMOS switch (usually 5 on the server's switch block) to resolve this issue. You will lose your server BIOS settings as they will be set back to the factory defaults.
Chapter 9 Issues and Problem Resolution controllers and associated containers into a single file. For example, if there are two controllers, then two commands must be issued, forcing the configuration of each controller to be saved into a separate file. Workaround: Use multiple commands to make multiple files for saving the multicontroller configuration. All of the separate files must be saved on the same diskette.
Chapter 9 Issues and Problem Resolution Problem: If a drive (e.g., drive 0:15:0) was assigned as a failover drive for a container (e.g., container 01) and then a second drive (e.g., drive 0:10:0) was assigned as a second failover drive for the same container, the second failover drive appears at the top of the assigned failover drives list. The assumption is that the first assigned drive should be the first on the list. Workaround: No action required.
Chapter 9 Issues and Problem Resolution • When the original controller is replaced with another controller. This situation would occur only if the user were planning to use the same containers and wishes to keep the data they contain. • When a new controller is installed in a system with preexisting containers. • A “controller rescan” command is issued to detect the new devices. The detected containers are not automatically exposed to the OS. That is, the containers are not accessible by the OS.
Chapter 9 Issues and Problem Resolution Entries in the Syslog are recorded as Errors, Warnings, or just Information. Since all activities of the controller fall into one of these categories, during normal operation, the entries in the log will be very limited. The number of entries increases significantly during failures and errors such as when Unix messages refer to the controller as AAC rather than HPN. Impact to Customer: Significant.
Chapter 9 Issues and Problem Resolution Problem: When one of the currently running services like CLI, SNMP agent or the AIF daemon dies, there is a possibility that it will not exit gracefully, thereby causing this problem to be triggered. In order to receive AIF information and display the data, a user program must register with the driver by calling an IOCTL. When the user program no longer wants to receive and display AIF information, it must unregister with the driver.
Chapter 9 Issues and Problem Resolution database is not actually damaged. Note that this problem only occurs if you are running version 5.5.2608.0 of the Ese.dll file. When Exchange Server makes a request to the hardware to read data from the database, it compares the returned data page and the calculated data checksum with the page number and checksum that were placed on the page when it was written. If a mismatch is detected, a -1018 error is generated.
Chapter 9 Issues and Problem Resolution Problem: One of the new features that was introduced in the latest code release was a Performance Enhancements for Windows 2000 Miniport driver. This feature allows a user to turn this option on and gain higher performance under Windows 2000. Since the Performance Enhancement disables some functionality, it is recommended during Benchmarking tests only. This feature can cause a blue screen when enabled.
Chapter 9 Issues and Problem Resolution Workaround: The issue has been fixed by changing Drive_Config to remove entries for product in mscsi file added by the OS during system boot and replace with entries for all supported product devices. 40 SCSI probe taking too long failing HCTs Problem: Microsoft HCT (Hardware Compatibility Test) sometimes reports an isalive failure caused by SCSI bus probe taking too long. The HCT test involves multiple reboots and blue screens of one of the systems for 24 hours.
Chapter 9 Issues and Problem Resolution Problem: Microsoft WHQL (Whistler HCT) is now testing the driver’s ability to address the entire 32 bits of memory. It has been determined that our communication region can only exist in the lower 2 GB of memory. Impact to Customer: Minimal, there is an error message in the event log: BAD_FW_WARNING Frequency of Occurrence: Often. Every time the system is rebooted, the entry is placed into the System Event Log.
Chapter 9 Issues and Problem Resolution Problem: When the Toptools Agent is installed on Windows NT 4.0 or Windows 2000 and a NetRAID-4M manageability agent is installed, but the management driver, located in the FAST utility kit, is not installed, the following event is generated by the SNMP service in the event log upon reboot: Event ID: 1106 Source: SNMP Type: Warning Category: None Resolution: The NetRAID-4M manageability agent will only be installed when the NetRAID-4M card is present.
Chapter 9 Issues and Problem Resolution Problem: Under Windows NT 4.0 and Windows 2000, the NetRAID-4M controller offers two types of formatting for containers: Full Format (with verify) and Quick Format (without verify). Neither Full Format nor Quick Format invokes verify. Full Format is the default; however, it takes longer than Quick Format due to the fairness algorithm introduced into Release 2. Workaround: For faster formatting, use the Quick Format checkbox.
Chapter 9 Issues and Problem Resolution Problem: If you run Windows NT 4.0 Setup from the three boot floppies, NT 4.0 fails to detect any logical drives or the container created with the NetRAID-4M during the text portion of NT 4.0 Setup, even if you load the NetRAID-4M driver. The root cause is the txtsetup.oem file on the driver's diskette for the NetRAID-4M. The workaround was to use the Navigator Installation Assistant, or boot from the Windows NT 4.
Chapter 9 Issues and Problem Resolution Resolution: Issue listed for information only. No action required. 52 Marathon not supported on Windows NT 4.0 Problem: The Marathon Technologies software architecture is geared towards working with the standard Microsoft Class/PortMiniport driver stack. In Windows NT 4.0, the NetRAID-4M has its own Class driver for RAID, which is incompatible with the current Marathon Technologies software.
Chapter 9 Issues and Problem Resolution Case 2: Under the Device Manager, even after specifying the A: drive, selecting the update driver option and allowing the utility to search for a driver will always pull the inf information from the c: drive that is already installed. We do not have the CAT file yet for the new release of the 4M Windows 2000 driver, and that is why the Microsoft OS cannot see the driver on the floppy disk.
Chapter 9 Issues and Problem Resolution This problem results in the following issues/system behaviors, as described in the cases below. If a specific workaround is not listed with the case, please refer to the tables below (6-1 and 6-2) to locate the NOS system behaviors that may occur and the appropriate tool to use for obtaining your desired results. General Example: You have a Windows 2000 basic disk. This happened either because it was migrated from Windows NT 4.
Chapter 9 Issues and Problem Resolution • Problem: During conversion from Windows 2000 basic to dynamic disk, a multi-boot partition will no longer be able to boot any NOS other than Windows 2000. • Background: A multi-boot partition is no longer able to boot to any NOS other than Windows 2000 because the Windows 2000 dynamic disk is a super-set file system.
Chapter 9 Issues and Problem Resolution Case 7: • Problem: If you use FAST to create Windows 2000 basic disk (HP NetRAID3Si does not allow this), and then files are put on it, and Disk Manager converts the disk to dynamic, all data is lost. • Workaround: This capability has been removed from the NetRAID-4M Windows 2000 tool. Use Windows 2000 Disk Manager for creation of the basic disks.
Chapter 9 Issues and Problem Resolution Table 9-2. Free Space and File System Volume Extension Free Space and Volume Extension Extend Disk Capacity Free Space Windows 2000 Dynamic Additional free space Disk Volume Created never appears until there From a New Empty Disk is a reboot. 1 2 Extend the File System Volume If basic disk, then you use DM2 to convert the disk to dynamic, then you create volume, then extension is OK. (This feature is not in the BIOS, CLI, or GUI).
Chapter 9 Issues and Problem Resolution Problem: If you initiate a manual move of resources in a Windows 2000 cluster, some of the resources will not move the first time. You must attempt the move a second time for them to move successfully. This only happens the first time that you manually move a resource. Any subsequent times, the Move Group function operates properly. Workaround: Repeat the move.
Chapter 9 Issues and Problem Resolution In Windows NT 4.0, a non-I/O pull push (removing and re-inserting a drive immediately), detected by SAF-TE, will create a yellow event in the Windows NT 4.0 system error log. The FAST GUI utility running on a Windows based client for a remote connection to NetWare will not be in sync, either on the server operating with NetWare or on a client console. In NetWare, a non-I/O loss, pull push will not create an event message on the system console screen.
Chapter 9 Issues and Problem Resolution arrays) get created, using one of the NSS Providers (Update Provider Information option), we can retrieve the newly created containers and show them as free space. If any of the new containers changes in size or gets deleted, the NSS Update Provider Information option does not reflect these changes, and neither does “List Devices” or “Scan devices for new hardware” options.
Chapter 9 Issues and Problem Resolution • NetRAID-4M installed • NetWare 4.2 or 5.1 • SymBIOS agent is loaded (symsnmp.nlm - Ver1.01d - 7/1/99) or if the SymBIOS SCSI-status agent is loaded (symtrap.nlm) • NetRAID-4M FAST utility is started on a remote client Workaround 1: Delete the Agents. Either rename or remove the file “sys:system\symsnmp.nlm” and “sys:system\symtrap.nlm” so it will not load the next time. Then reboot the LP 2000r.
Chapter 9 Issues and Problem Resolution Background: UnixWare 7.1.1 comes with a feature called MPIO (multi-processing IO), a feature similar to clustering where two controllers can access the same hard disk. MPIO is not installed by default but can be easily selected. MPIO seems to be interfering with the boot process. Workaround: Re-install UnixWare without MPIO.
Chapter 9 Issues and Problem Resolution 3. Within the DCU, use the Hardware Device Configuration to ensure that the hpn driver is mapped to the hpncard. If the hpnraid driver is mapped in the second column to a device with resources attached to it, as indicated in the other columns to the right, the driver has most likely mapped correctly.
Chapter 9 Issues and Problem Resolution These error messages usually occur when the entire free space has been allocated for partitioning or after using DISKADD, and these error messages appear with or without the NetRAID-4M controller installed. Workaround: Ignore the errors. No action is required. 68 FSCK failure Problem: After power down and reboot, file system check (FSCK) will sometimes fail with a system panic during boot.
Chapter 9 Issues and Problem Resolution ¾ HP NetRAID o From the BIOS Main Menu select Embedded LAN and SCSI settings. Disable “Embedded SE SCSI.” 2. Disable the embedded Symbios SCSI controller by pressing Ctrl + C during boot up: o Select “Change Adapter Status” and press Enter. Change both of the “NextBoot” settings for the SYM53C806's from ON to OFF. 3. Exit the utility and restart. 70 OpenServer 5.0.x reports “Spurious Interrupt” Problem: OpenServer 5.0.
Chapter 9 Issues and Problem Resolution When you are ready to set up your disks under OpenServer, do not use mkdev hd. You must first use fdisk -f and then divvy -m to configure your 4M disks. For example, fdisk -f /dev/rdsk/c0t0d0 Use the CLI utility to open the 4M controller and find the device names such as c0t0d0 for each created container. For example, to launch the CLI and list all the containers: /usr/sbin/.
10 Glossary Agents In general, every agent monitors a specific entity. Microsoft provides an agent that monitors the data specified in SNMP standards documents. This information includes network performance and configuration data. Almost every other agent monitors some piece of proprietary data. Agents, OEM The HP Server agents include a number of OEM-provided agents that add instrumentation and monitoring for the storage subsystem.
Chapter 10 Glossary Cluster (In A File System) A cluster is a group of contiguous sectors and the smallest increment of storage. The cluster size is determined by the total partition storage space divided by the number of addressable clusters. For example, a 2 GB partition divided by 216 addressable clusters would result in a cluster size of 32 KB (64 sectors). Cluster (In High Availability) A cluster is a unit of computing resource that can include, for example, storage and IP addresses.
Chapter 10 Glossary Disk Manager (DM) MS Windows 2000 tool. It is equivalent to the NT4 Disk Administrator. Drive Letters Drive letters are assigned to volumes for easy identification when referring to the exact location of a file. Drives per Container: Table 10-1.
Chapter 10 Glossary File System Expansion A term used to describe extending a file system (a feature of Windows 2000), such as NTFS. In NT 4.0, no volume extension was needed to do this. In Windows 2000, it appears that volume extension, when NTFS is on a volume, is the only way to extend a file system. Free Space The HP NetRAID-4M works on the concept of “free space” rather than dedicated spare disks.
Chapter 10 Glossary Manageability Describes the various activities that need to be done to keep a computer network up and running. Activities include inventory collection, software distribution, alert notification, as well as others. Toptools provides a useful subset of all the manageability tools needed to effectively manage a network. Novell Storage System (NSS) Allows the combination of free space from many areas into one large, logical volume.
Chapter 10 Glossary Partitions - System vs. Boot Microsoft uses the term “system partition” to refer to the partition that contains the boot files such as NTLDR, and the term “boot partition” to refer to the volume containing the Windows NT 4.0 operating system files. This terminology is reversed from general industry usage. RAID bad block replacement (BBR) A process automatically invoked when a controller cannot read or write to the media. BBR attempts the write three times to the same block.
Chapter 10 Glossary RAID containers (NetRAID-4M) Containers are logical disks (also known as arrays) created from free space and made up of one or more partitions on one or more physical disks. A container that spans multiple physical disks can be larger than any one of the physical disks. Containers differ from most other RAID arrays in that their underlying partitions can be smaller than a physical disk. Consequently, several containers' partitions can reside on a single physical disk.
Chapter 10 Glossary In the horizontal dimension, the RAID (even) parity information detects an error. In the vertical dimension, the drive read/write electronics also detects an error. Together, the error is located and corrected. status parity Drive H Drive G Drive F Drive E Drive D Drive C Drive B Drive A Table 10-1.
Chapter 10 Glossary RAID level migration feature 1. Sometimes referred to as container reconfiguration and container expansion. The ability to change the RAID and/or add disk drives to a RAID without having to back up and restore the data. NOTE You cannot migrate to or from a RAID container that is using more than 32 drives (even though the NetRAID-4M can contain up to 48 drives). NOTE RAID, Multilevel Container A multilevel container is a combination of one or more single-level containers.
Chapter 10 Glossary Table 10-2. RAID-10 examples Drives Total RAID-1’s Total available Total RAID-10 Total hidden in a single per drives containers containers being used RAID-1 RAID-10 2 2 1 24 24 (1 per RAID-10) 4 2 2 20 40 (2 per RAID-10) 8 2 4 10 40 (4 per RAID-10) 12 2 6 6 36 (6 per RAID-10) 24 2 12 3 36 (12 per RAID-10) 36 2 18 2 36 (18 per RAID-10) 40 2 20 2 40 (20 per RAID-10) 48 2 24 1 24 (24 per RAID-10) Table 10-3.
Chapter 10 Glossary RAID-10, Multiple 48 Drive Containers The controller design supports a total of 64 containers of which 24 are accessible to the user and 40 are hidden. The first 48 drive RAID-10 uses up 24 accessible and 24 hidden. Therefore, there are not enough leftover containers to make another RAID10 in the same HP Server. RAID member fault An example of a RAID member fault is a failure of the first drive of a not-faulttolerant container (volume set or stripe set).
Chapter 10 Glossary Duplex (simple) - A special case of RAID-1 where the drives are on separate channels (same controller) for increased fault tolerance. Duplex (teaming) - Another special case where the drives are on separate channels that happen to be on separate controllers. Booting from this type of duplex is technologically difficult because it requires that the system BIOS know which pair of the card and mirror half is good and which pair is bad.
Chapter 10 Glossary RAID zero-channel A zero-channel RAID is a PCI card with an input output processor (IOP) but without the SCSI ASIC. The zero channel RAID card fits into a special PCI socket whose reserved pins have been wired directly to the SCSI controller that resides on the motherboard. SAF-TE - SCSI Accessed Fault-Tolerant Enclosures SAF-TE is a standard developed by nStor Technologies and Intel. The enclosure itself is treated as another SCSI device.
Chapter 10 Glossary from anywhere on the LAN using a web browser like Internet Explorer. Toptools requires instrumentation on each managed device/target. Instrumentation provides information on warnings, faults, temperature, inventory, performance, etc. For HP Servers, the instrumentation is provided by the HP Server Agents. Toptools, Instant Instant Toptools is a reactive support tool that installs on a single HP Server.
Chapter 10 Glossary 89
Chapter 10 90 Glossary
Index A AAC entries 50 Active Directory 60 agent 23, 24, 52, 56, 68, 75, 87, 88 Automatic Failover 39, 56 B background consistency check (BCC) 22, 44, 67, 80 bug fixes 7 build numbers by release 9 C Caldera support 6 cluster 3, 11, 50, 66, 76 CMOS 47 Command Line Interface (CLI) 1, 7, 8, 9, 11, 12, 14, 52, 54, 65, 68, 70 common tray cabling 29 components firmware, drivers, utilities 7 91
container boot 70 configuration 48 creation 43, 48, 59, 67, 68, 73 critical 81 definition 76 degraded 44 deletion 42 expansion 43 failover drive 49 foreign 50 formatting 58 logical drive 78 multiple 48 OpenServer 73 RAID-5 66 redundant state 56 scrubbing 86 size 39 size change 68 striping 58 UnixWare 69, 71 unprotected 86 Container Configuration Utility (CCU) 43, 47, 48, 66, 69 controller 3, 8, 10, 47 controller rescan command 50 D data corruption 39 disk basic 61, 62, 63, 64, 75 dynamic 79 DISKADD 72 E e
error(s) 1018 drive 52 BAD_FW_WARNING 56 bogus 57 coding 60 consistency check 80 container size 39, 43 DISKADD 72 drive 80 duplicate entries 54 extending file system volume 64 isolation 42 NetWare installation 67 NetWare, remotely managed 67 network 39 no data 46 rebuild onto failing drive 56 SCSI bus parity 46 stack 40 Syslog 51 timeout 49, 52 Ultra3 hard disk drives 23 uncorrectable read 57 unrecoverable 48, 78 F H failover 39, 43, 48, 52, 56, 78 failover drive (hot spare) 48, 49 fault 48, 52, 67 file
ID device 70 event 57 SCSI 3, 39, 62, 78 vendor 70 internal cabling 29 K L kernel 42, 51, 52, 73 Linux 5, 12, 14, 15, 50, 51, 52 Linux support 5 logical drive 78 M manual installation 67 manual override 3 Marathon 55 member fault 85 miniport driver 10, 54 mirroring 76, 77, 80, 83, 85, 87 N O P 94 Navigator 24, 25, 57, 58, 67 NetWare 5, 12, 14, 15, 66, 67, 69 OpenServer 12, 14, 15, 50 operating systems supported 5 panic 42, 43, 52, 72
partition 88 active 79 as volume 88 boot 80 cluster size 76 dead 81 extended 79 formatted 79 full 88 in containers 76, 81 logical 88 primary 79 rebuild 85 striping 87 system 80 table and signature 75 versus "volume" 79 R rack storage 27 RAID-0 58, 77, 85 RAID-1 69, 77, 84, 85 RAID-10 77, 84, 85, 86 RAID-5 76, 77, 83, 84, 85, 86 RAID-50 77, 84, 86 rebuild 85 Automatic Failover 38 degraded performance 81 failover drive 49 failure 42, 81 fast formatting 58 free space 48, 78, 82 kernel 52, 73 scrubbing 86 RS/1
SCSI termination 3 server environments supported 17 slotting requirements 17 software component downloads 10 software components 7 striping 58, 76, 77, 80, 81, 83, 85, 86, 87, 88 Syslog 50 system crash 55 system support 5 U Ultra3 hard disk drives firmware issue 23 UnixWare 5, 10, 12, 14, 15, 35, 50, 69 unprotected 86 updating 11 upgrading 10 V volume extension 62, 64 free space 67 set 61, 76, 88 space 62 spanning 88 W 96 Windows 2000 15, 39, 54, 55, 57, 58 Windows NT 15
