IBM System Storage™ DS4000 storage subsystem controller firmware version 06.60.34.00 for the DS4200-all models, DS4300 (dual-controller Standard/Base and Turbo models only), the DS4500-all models, the DS4700-all models and the DS4800-all models storage subsystems only This firmware version is required to support the attachment of the DS4000 EXP810 storage expansion enclosures to the IBM DS4000 TotalStorage™ DS4300 dual-controller Standard/Base and Turbo, the DS4500, and the DS4800 storage subsystems. It is also required to support the intermix of the EXP810, EXP710, and the EXP100 storage expansion enclosures behind the IBM DS4000 TotalStorage™ DS4300 dual- controller Standard/Base and Turbo, the DS4500, the DS4700 and the DS4800 storage subsystems. The IBM DS Storage Manager host software version 10.60.x5.17 or later is required to manage DS4000 storage subsystems with controller firmware version 6.60.34.00 installed. Last Update: 11/16/2010 Version History ----------------- This is an updated release of IBM DS4000 System Storage™ controller firmware/NVSRAM in the 6.60.xx.xx controller firmware code thread for the DS4200-all models, DS4300 (dual-controller Standard/Base and Turbo models only), the DS4500-all models, the DS4700-all models and the DS4800-all models storage subsystems only. October 06, 2010 - version 06.60.34.00 - Updated NVSRAM to: + Fail drives upon certain thresholds for stagnant I/Os (Cal 6105). Severity 1 - Fix 165738 (CL #WIC~151915) False PFA for Stagnant I/O failures reported in MEL if multiple lUNs from a single array are owned by both controllers. - Fix 178553 PANICs during controller reboot stress test. - Fix 177842 Data abort seen after entering VKI_EDIT_OPTIONS statement. - Fix 165114 Ancient I/O seen during 30 hour sysReboot stress. - Fix 164271 (cXB015476) VMware data corruption when write command aborted. - Fix 158752 (cXB010862) ECC errors on one controller cause outage to the alternate controller. Severity 2 - Fix 181151 (RW #181024) (CL #WIC~169536) Collecting ddc "on demand" doesn't log task info. - Fix 181024 (RW #180352) (CL #WIC~169536) Collecting ddc "on demand" doesn't log task info. - Fix 180352 (RW #179834) (CL #WIC~169536) Collecting ddc "on demand" doesn't log task info. - Fix 150008 (c448218) Volumecopy corruption after target expansion. - Fix 179834 (CL #WIC~169536) Collecting ddc "on demand" doesn't log task info. - Fix 177416 (cXB032755) PCI SERR Exception during parity scan activity. - Fix 178367 watchdogTest shell command results in an exception. - Fix 178521 (CL #WIC~154365) Drive migration yielded Incompatible drive. - Fix 177000 Automated test failed with host loss of communication during failed drive reconstruction. - Fix 154542 Incorrect drive states after insertion of drives. - Fix 156850 (CL #WIC~156798) Inband getChangeState requests fill all UTM structures. - Fix 159928 (c88378,004) Global fault LED stuck on in a 4600 tray when system is optimal. - Fix 160575 (cXB013311) Watchdog timeout while gathering support bundle. - Fix 163590 (CL #WIC~163093) Leverage the DDC collection mechanisms to provide on demand DQ collection tool. - Fix 164904 "Out of memory" returned on shell commands during Support Bundle collection. - Fix 167851 (cXB010991) Disable Watchdog Timeout DDC Trigger by Default. - Fix 173480 (CL #WIC~171020) Network Storms are causing controllers to reboot. - Fix 163380 (cXB014546) Repeated CTRL reboots due to WTO - “IOSched and IOSched2 are waiting to allocate buf_t’s”. - Fix 154196 (c463787) Secondary Mirrored Volumes does not receive IO, Status 'Unknown’. Severity 3 - Fix 173388 (CL #WIC~171866) getCtlSocErrStats uses static data struct, to gets -> panic. - Fix 173460 (cXB029286) MEL logs volume as 'copy complete' in a copyback process which never occurred. March 05, 2009 - version 06.60.22.00 -Fix 152946 (CL #WIC~112299) CLY - Assertion failed: overflowDetected file spmSPMMgr.cc line1013. -Fix 152980 NVSRAM mismatch on 399x/6091 config after warm controller module replacement. -Fix 153045 Media Scan does not report 2033 Parity Repair operation when media scan performs parity repair. -Fix 149450 (CL #WIC~140998) (c421316) Panic/Reboot loops with very large RAID-1 at vdmRAIDImplExclOp.cc. -Fix 150869 (CL #WIC~147022) Drives were lost after channel instability during a sysReboot. -Fix 151010 NVSRAM mismatch at 5884 array after warm controller module replacement. -Fix 152455 "Redundancy Group Not Consistent" is not logged in MEL. -Fix 152461 Dual controller reboots at PANIC: VKI_KMFREE - Null block pointer. -Fix 152592 (CL #WIC~134551) Controller panics after upgrade from 6.17 to 6.30 or later if persistent resv used. -Fix 124384 (CL #WIC~124293) Unrecovered Read Error Returning LBA of drive. -Fix 136592 (c409117) Alarm does not go off after power supply is replaced. -Fix 136755 Removing fan A causes a fan B failure status. -Fix 140534 Extraneous 101F MEL events occur after media error drive PFA set. -Fix 142488 (c423901) Date of manuf for drives missing in Profile. -Fix 145844 The 2036 Mel event does not match MEL specification. -Fix 147889 (c443093) Exception: Divide Error. -Fix 131533 (CL #WIC~131472) Encountered Ethernet register read error during power-up diagnostics. -Fix 131608 (CL #WIC~126125) Many PFA (Stagnant I/O) failures reported during I/O test run. -Fix 140456 Include host WWN in raw data of logged AVT MEL event. -Fix 143725 (c435864) PCI_LOCKDOWN entry in the DDC trace log may be incorrect. -Fix 143743 Global fault LED stuck on in a 4600 tray when system is optimal. -Fix 143807 (c431401) After upgrading firmware, the controller asserted as vdSpecialOp.c, line 1240, mappings lost. -Fix 143958 (CL #WIC~136048) Mapsets between the two controllers don't match. -Fix 144409 (CL #WIC~143783) ensure volume configuration is saved correctly after copyback. -Fix 145836 (c435339) Controller replacement causes SPM database corruption. -Fix 146895 (c424879) Flashcopy problem on a mirror secondary logical drive. -Fix 147003 Controller reset during volume build under heavy load. -Fix 147868 (c443192) Old host information not cleaned up when PLOGI processed. -Fix 148138 (CL #WIC~146502) SLES10 Hosts fail to discover all volumes post-UTM. -Fix 148640 (cXB003540) RVM deletion operation staying queued and causing loss-of-management. -Fix 149095 (CL #WIC~142218) 1 TB SATA drive will not start volume reconstruction. -Fix 149097 (CL #WIC~138673) Controller reads data from a "Piece Replace. -Fix 149098 (CL #WIC~143569) Performance enhancement for initializing 1T. -Fix 149099 (CL #WIC~145727) Drive becomes optimal after skipping recons. -Fix 146946 (CL #WIC~141996) Abnormal packets sent from the management ports on the controller. -Fix 143952 (CL #WIC~138044) (c411708) scsi2 reservation not being passed to hosts. September 05, 2008 - version 06.60.17.00 - NVSRAM file updates. -Fix Solaris Host Region 8 default NVSRAM setting changed to support >32 luns. -Fix 140691-Disable Parity Repair during Parity Scan. Parity Scan default setting changed to Data/Parity mismatches encountered during parity scans will NOT be corrected. -Fix 143377 RPC Timed out and caused loss of connection from array to Storage Manager. -Fix 142336 (RW #133812) Overlapping getDriveLogData commands cause Ancient I/O -Fix 142808 Mirror Repository size is .126GB instead of 128MB -Fix 143483 (RW #134136) (CL #WIC~133906) 200A MEL event data ID corrupted -Fix 138608 (c415528) Invalid scrub_flags value triggered continuous panic assertion reboots -Fix 139832 (c417549) The recreate Snapshot command creates a 0 capacity Snapshot -Fix 140164 Ancient I/O because of deadlocked controllers -Fix 141106 SYMbol upgrade of CHR2M2 NVSRAM failed with Ctl A being held in reset -Fix 141926 (CL #WIC~141674) (CR6702775) Cannot add OEM system to Storage Manager -Fix 127371 Shell command component of DDC data is not present -Fix 135511 (c404478) Recover Command Fails on 2 LUNs within a system -Fix 125966 Controller stops responding to IP traffic after rlogin enabled -Fix 126483 Alarm Sounding After Being Disabled via MAPI Command -Fix 128682 Drive channel 1 shows up 3 times and 2 times as failed incorrectly -Fix 130794 (c393515) Drive group marked offline but no event was logged to the MEL -Fix 132701 RMVM Debug Library Cannot Register an Unmarshall Function -Fix 132792 (CL #WIC~132396) RVM mirrors failing during port -Fix 133674 (CL #WIC~133560) IO errors on Solaris MPXIO caused by old srcOps -Fix 133812 Overlapping getDriveLogData commands cause Ancient IO -Fix 134136 (CL #WIC~133906) 200A MEL event data ID corrupted -Fix 134157 (CL #WIC~134020) Multiple redundant DEGRADED_SYNCHRONIZING events being posted -Fix 134228 (CL #WIC~133964) Adding a command to dump information for MEL 200a Event -Fix 134368 VerifyBlocks causes assertion failure -Fix 134827 (CL #WIC~126544) (CR6572039) SYMbol is returning 0 for the 5884 Ethernet speed -Fix 135365 RMVM Debug Library takes memory fault -Fix 135640 (CL #WIC~130525) After a forced AVT, Ctl B went into a Panic Loop with “Invalid Cache Reclaim State” -Fix 136410 Assertion failed & controller went into Debug during sysReboot test -Fix 136860 (CL #WIC~124737) Controllers go into debug after removing host mappings and rebooting the controller -Fix 137178 Interoperability issue between the Tachyon and the SOC chip in the drive trays -Fix 138085 (CL #WIC~135443) (c400006) Stop Snapshot CMD Does Not Free Capacity in Repository -Fix 138620 (RW #132978) (CL #WIC~132581) Inquiry command to page 0xC1 sent to OEM SATA drives -Fix 138906 (c406730) (CL #WIC~135369) Corrupted Snapshot repository caused controllers to continuously panic -Fix 138907 (CL #WIC~137130) Failover occurs due to a log sense defect -Fix 138363 Unable to upgrade from 06.19 OEM builds to 06.60 OEM builds -Fix 137376 Unable to upgrade from 06.19 OEM builds to 06.60 OEM builds -Fix 136092 (CL #WIC~135646) Performance improvements for persistent reservations -Fix 142301 (RP #WIC~141106) SYMbol upgrade of CHR2M2 NVSRAM failed with Ctl A being held in reset -Fix 136227 (c396194) execLogShow CMD Proc Exception -Fix 137483 Clearing the FA log without reboot causes the entries not to show up in excLogShow output March 04, 2008 - version 06.60.08.00 - DS4300 NVSRAM files corrected. The incorrect DS4300 NVSRAM file versions were referenced in the previous v06.60.08.00 readme file dated Feb. 21st, 2008. In addition, the incorrect DS4300 NVSRAM file versions were placed in the DS4300_FAStT600_Base and DS4300_FAStT600_Turbo subfolders; For_attachment_of_FC_enclosures_Or_FC-SATA_intermix_only and For_attachment_of_SATA_enclosures_only. The correct DS4300 NVSRAM files are now; N1722F60R960V08F - For_attachment_of_FC_enclosures_ Or_FC-SATA_intermix_only N1722F60R960V08S - For_attachment_of_SATA_enclosures_only Note a suffix has been added to the end of the DS4300 NVSRAM file to help identify the correct file. Feb 21, 2008 - version 06.60.08.00 - NVSRAM change: Changed AIX, Host 4 User Config AIX-ADT/AVT and Host 6 User Config AIX, Retain Logins low order bit of Host Options 1 set to 0. - Fix 124503 (c376104) Daily Multiple Watchdog Timeouts occurring after CFW upgrade. - Fix 129429 (c388614) RVM re-activation fails if full capacity is used. - Fix 133085 (c398724) (CL #WIC~131136) Drive enclosure fault light still on after disk replacement. - Fix 132479 DS4800 cache suspended 06.23.05.xx code level. - Fix 128953 (c387144) Storage Manager Drive Channel Report is bogus. - Fix 129129 (c388324) Unable to sync RVM due to Primary system displays mirrored secondary LUN as zero GB. - Fix 132116 (c386847) (CL #WIC~123891) GUI unresponsive after RVM was established. - Fix 126699 (c369549f_1) (CL #WIC~106870) DDC was triggered before the specified threshold and it should not be triggered in that wind. - Fix 128144 (c385071) Controller failed to lockdown due to Memory ECC Errors. - Fix 128164 (c385522) Controller seeing Watchdog timeouts after code upgrade. - Fix 128856 (CL #WIC~121079) (DR112895) Degraded drive-side channels reported when they have not exceed default error threshold - Fix 128958 (CL #WIC~119137) "Not enough memory" message when "hotspare coverage" is selected when a hotspare is in use. - Fix 129197 (CL #WIC~126241) Watchdog: Controllers debuged due to a management network Broadcast storm on network management port of the controllers. - Fix 129485 (c383362) Event ID 6900 missing from the MEL though Recovery Guru is reporting the DDC data is available to be collected. - Fix 129768 (CL #WIC~113451) (CL #WIC~112993) Unused SFPs in drive enclosure show as failed in the Recovery Guru when it is not. - Fix 129865 (c353454) (CL #WIC~128403) PANIC: DNM: Devnum assignment out-of-range, chan:15 id:0 when changes are made to subsystem drive-side. - Fix 129917 ECC error indication lost due to forced AVT. - Fix 129919 (CL #WIC~126314) Drive shows up with a preferred and Redundant Channel of 0 in Profile and State Capture. - Fix 129927 (CL #WIC~103689) Event specific code field in MEL returns incorrect ASC/ASCQ for drive check condition. - Fix 130852 (c384637/c481150) (CL #WIC~130207) Ancient I/O on High Stress SATA Only Configuration. Oct 09, 2007 - Version 06.60.02.00 - Recreate multiple snapshots. SYMbol addition of a re-create snapshot command that will accept a list of snapshots as input. This command will be used by VSS and SMI-S providers to reduce the time needed to re-create several snapshots at one time. - Fix 127685 (CL121820)(c368940) Simultaneous DDM/Controller failure - Fix 127196 (CL126235)(2-183925398) Ethernet Connection goes down - Fix 122511 (CL115612) (ST960048) - TPSSM 9.16 client reporting incorrect WWW - Fix 125632 (CL124379) (CL124071) (2-178252461) Bad Zone - Fix 124381 (CL124293) Unrecovered Read Error Returning LBA of drive - Fix 121533 (CL118110) (2-169210544) Trying to create a new Lun and parity - Fix 121537 (CL121285) (c362864) Controllerts rebooting due to SPM Corruption - Fix 121539 (CL 121285) (c362864) Controllerts rebooting due to SPM Corruption - Fix 122850 (c370875) After Changing LUN mapping; LUN is no longer recognized - Fix 106193 (D25918) WP/Wrigley error in the array profile for NL Port ID - Fix 118448 (RW #115398) (C356253) Intermittant failures, CCDL on DS4500 - Fix 122015 (CL117664) (+49 6103752647) fcCheckTimeElapsed Lost management - Fix 118544 (CL115239) (2-159282581) Configuration was corrupted - Fix 120530 (C367316) Incorrect logical drive name identified in one of the MEL entries - Fix 120582 (CL120350) Controller reports ASC/Q 25/00 to host on existing LUN - Fix 120641 Cannot enable audible alarm via GUI or script. - Fix 120643 (CL113124) (CR6491533) After an upgrade from 06.12.27.10 to 06.19.xx.xx July 05, 2007 - Version 06.23.05.00 readme update only - Updated the minimum MAX3036FD, MAX3073FD, and MAX3147FD drive firmware files to version S708. This firmware has the fix to the drive reporting an invalid frame as an 04/44/98 internal hardware error rather than a 0b/44/98 retryable error. April 26, 2007 - Version 06.23.05.00 - Support new DS4000 System Storage 4Gbps/15Krpm 300GB Fiber Channel Enhanced-Disk Drive Module (E-DDM) option - Support new DS4000 System Storage 750GB Serial ATA E-DDM option - Support the intermixing of the Fibre Channel and SATA E-DDMs within the same DS4700 storage subsystem or EXP810 drive expansion enclosure. - Support the intermix of EXP810/EXP710 and EXP100 enclosures behind a DS4700 and DS4800 storage subsystem. - Support new "Enhanced Disk Drive Predictive Fault Analysis" diagnostic feature on the DS4300, DS4500, DS4700 and DS4800 storage subsystem. The DS4000 controller will monitor and flag any the Fibre Channel and SATA hard drives in the DS4000 storage subsystem configuration with PFA error when one of the following conditions occurs a. 500 recoverable errors within a 24 hours period for Fibre Channel drives b. 4 medium errors within within a 24 hours period for Fibre Channel drives c. 8 medium errors within within a 24 hours period for SATA drives - Incorporate the following field defect fixes - Fix 113065 Unreported Asynchronous Event Notification (AEN) event is logged during controller reboot. Fixed by setting AEN event to be reported in eelClearEvent so it will not be re-posted with controller reboot. (CMVC 355313). - Fix 111471 Controller assertion failure: sstorStableStorageDrive.c line 73. Fixed by adding an additional check before asserting for a removed optimal drive. (CMVC 348072). - Fix 109408 Volumes not sorted in contiguous order after sorting by "Logical Drive Capacity" (CMVC 339565). - Fix 112538 (110822) Controllers don't agree with drive state and prevented reconstruction (CMVC 338559). - Fix 114169 Clear flags-2 bit in SCSI OP caused Sense data Unit Attention 06/29/00 to be sent to alternate controller (CMVC 308285). - Fix 115014 (113832) DS4700/DS4800 4Gb Data Negotiation Failure QLogic QLA2340 HBA (CMVC 339741). - Fix 115709 Firmware performance improvement for flashcopy recreate time (CMVC 354503). - Fix 116475 (115944) Persistent Reservation In (PRIN) was modified to use the shared transaction lock and to be batch processed so the lock overhead is encountered only once for all PRIN requests queued at that time. Moved a check for the presence of a SCSI-2 reservation on the alternate controller so that it only occurs if there is not a persistent reservation since Persistent Reservation Out (PROUT) enforces the mutual exclusion and PRIN can rely upon it to reduce its work load. Modified SCSI-2 reserve to only get the transaction lock if it is actually making changes. - Fix 116481 (97922) Cache enabled but not active. Corrects a problem where a controller reboots while its alternate is in Start of Day (SOD) also. In this case, the controller that reboots does not get a check-in message from the other controller and disables caching if mirroring is requested. The controller that did not reboot completes SOD and enables caching and mirroring since the alternate controller checked in during its second SOD sequence. CMVC 358385). - Fix 116973 (116311) Re-establishing Remote Volume Mirror (RVM) fails with error 176 after upgrading from controller firmware 06.12 to controller firmware versions 06.19.19.00 or 06.19.20.00. Error 176 was reported on mirror deletions and the mirrors would not synchronize (CMVC 358706). - Fix 117170 (116631) Using Async Mirroring with write cache enabled caused unexpected controller reboot. The firmware was modified to set a flag in the remote mirror proxy when I/O is being suspended and not allow the background synchronization to restart when this flag is set. (CMVC 340268) - Fix CMVC 368914 extra characters in host region name LINUXCLVMWARE for DS4200 (N1814D420R923V08) and DS4800 NVSRAM (N1815D480R923V09).