HP StorageWorks 9100 Extreme Data Storage System release notes V1.0.
Legal and notice information © Copyright 2009 Hewlett-Packard Development Company, L.P.
Purpose These release notes apply to HP StorageWorks 9100 Extreme Data Storage System version 1.0 and later. Version identification To identify the version of the ExDS system software use the following command: # exdsmgr show version The release and version strings are described in the following table: Release Version number V1.0.2 V1.0.2–4811 V1.0.1 V1.0.1–4310 V1.0 V1.0–3972 Update recommendation Recommended Supersedes HP StorageWorks 9100 Extreme Data Storage System versions 1.0 and 1.0.
Contents of the 1.0.1 release Firmware version change summary The 1.0.1 release of HP StorageWorks 9100 Extreme Data Storage System contains updates to the firmware of the following components: • iLO: new version: 1.70—This is the minimum supported version. The integrated Lights Out (iLO) is automatically updated as part of the update process. • ExDS9100cc: new version: 1.26.2008120502—This is the minimum supported version. The process is described in the update instructions.
SAS switch module management firmware 1.1.8.0 changes since version 1.1.6.0 • Fix for possible port map issue that may disable a data path between a server blade and the SAS switch • Fix to firmware upgrade process to avoid a potential upgrade lockup issue Added an exdsmgr command to delete the alert history This command allows you to purge the alert history. The syntax is: delete alert history ||all The exds_stdiag command now shows the percent of parity initialization of a degraded LUN.
• 25166: When the syscheck command reported a degraded LUN, it identified all arrays as the location of the LUN. This has been corrected. • 25282: When an ExDS system is powered on from a fully down state, the system required a reboot of one or more blade servers. The system now boots to an operational state without manual intervention. • 25295: The daily purge process for /var was broken. This is now fixed. • 25330: In some situations, a server blade can crash with psfs_mpbbuf.c:254 assertion failure.
This command takes up to 20 minutes to update the software on all servers. During the update process, the servers reboot. At the end, the server where you are running the update procedure will reboot. 5. After the server where you are running the update procedure has rebooted, complete the following steps: a. Log into the server again. b. Execute the update script a second time to complete the update: # cd ExDS-1.0.2-4811_update # ./InstallUpdate1.0.2.
6. If you are updating from an HP ExDS9100 1.0 system, you must update the firmware now. Follow steps 7 through 16 below. If you are updating from an HP ExDS9100 1.0.1 system, no further action is required. 7. Shut down the system so that only the first server is running in single user mode as follows. This process is described in chapter 5 of the HP StorageWorks Extreme Data Storage System 9100 administration guide. • If the system exports filesystems using NFS, disable all export groups.
f. 9. As the controllers power on, the 7-segment display on all ExDS9100c enclosures displays a number of states and numbers. This is expected behavior. When the 7-segment display shows on, the ExDS9100c enclosure is operational. Do not interrupt the process. Wait up to 10 minutes to allow all 7-segment displays to show on. If at this stage one or more storage blocks show any code other than on, power cycle the entire storage block (that is, both the ExDS9100c and ExDS9100cx).
NOTE: Sometimes the update process fails. If this happens, reset the SAS interconnect and repeat the firmware update process. This is a known problem that should not occur once the SAS switch module management firmware is updated to this revision. f. g. Disconnect from the interconnect by pressing Ctrl+Shift+_ (the control, Shift, and underscore keys) simultaneously. Repeat the process for interconnect 4 (that is, the second SAS switch module). 14.
Issue Description Sometimes a fan on an ExDS9100cx chassis reports a spurious DEGRADED status. 24591 The status is reported in a log message similar to: SGA8320003: array 3: box 2: fan: 2: fan is DEGRADED (was OK). When the status returns to OK, a message such as this is logged: SGA8320003: array 3: box 2: fan: 2: fan is OK (was DEGRADED). Workaround: If the OK message is reported within a few minutes of the DEGRADED message, you can ignore this status as spurious.
Issue Description The green status/activity LEDs on certain drives may not be lit correctly to show the drive state, and may intermittently revert to an unlit state . 25243 In certain situations the green status/activity LEDs may not be lit correctly to show the drive state, and may intermittently revert to an unlit state although the drive is configured and part of a provisioned LUN. Workaround: This issue does not affect the drive functionality, nor does it affect the amber fault LED indicator.
Issue Description The exdsmgr command sometimes responds with: "Unable to contact database...The service may be transitioning. Please try again." This message can indicate that a server has failed. Within a few seconds the database should failover to a backup server and the exdsmgr command will work normally. However, if the problem persists, there is a problem with access to the database. Workaround: You can isolate the problem as follows: 1.
Issue Description The exdsmgr install update command requires full path names for the RPM files. Relative path names do not work. 26253 The error message is not very clear; when it cannot find the RPM, it displays a messages such as Error: Invalid flag (example.rpm) entered. Workaround: Specify full path names. In some configurations, power supply 3 and 6 are not connected to a power source. 26339 The exdsmgr syscheck command correctly ignoresthis situation.
Issue Description HP ExDS9100cc controller has amber LED and H1 or H2 code in the 7-segment display. If and ExDS9100cc controller has an amber LED and the 7-segment display shows a code of H1 or H2, this could indicate a failure of one of the ExDS9100cx I/O modules. The H1 or H2 code means that the controller (1 or 2) has been forced to shut down by its peer controller. Workaround: Check the GSI failure LED on the front of the ExDS9100cx enclosure.
