Patch-ID# 113722-11 NOTE: *********************************************************************** READ THE TERMS OF THE AGREEMENT ("AGREEMENT") IN THE LEGAL_LICENSE.TXT FILE CAREFULLY BEFORE USING THIS SOFTWARE. BY USING THE SOFTWARE, YOU AGREE TO THE TERMS OF THIS AGREEMENT. IF YOU DO NOT AGREE TO ALL OF THE TERMS, PROMPTLY DESTROY THE UNUSED SOFTWARE. *********************************************************************** Keywords: storedge 3310 se3310 se3000 raid storage firmware safte security Synopsis: SE3310 413B: StorEdge 3310 array controller firmware upgrade Date: Aug/26/2005 Install Requirements: NA Solaris Release: 8 9 10 SunOS Release: 5.8 5.9 5.10 Unbundled Product: StorEdge 3310 Array Controller Firmware Unbundled Release: 413B Xref: Topic: Relevant Architectures: sparc i386 NOTE: If you are using 3.25W or earlier controller firmware and do not want to upgrade to version 4.13, you can download the most recent 3.25 firmware patch, 113722-09, from sunsolve.sun.com. NOTE: If the StorEdge PCI Dual Ultra3 SCSI Host Adapter, x6758A, is used to connect the StorEdge 3310 array, the qus driver must also be updated. The updated qus driver is available in the following patches: Solaris 8 Operating Environment: 112697-04 Solaris 9 Operating Environment: 112706-03 NOTE: When upgrading from 3.xx to 4.13 on Solaris, Windows, or Linux, you must use the upgrade script for 3.xx to 4.xx upgrades. Other platforms require the manual upgrade procedure contained in this README file. * See the quick_upgrade.pdf file contained in this patch. It is for experienced Sun customers who can use the upgrade script and minimum instructions. * See the script_troubleshooting.pdf file contained in this patch for script troubleshooting information. BugId's fixed with this patch: 4702532 4714227 4714811 4717055 4722548 4723464 4726561 4739050 4743721 4764506 4792748 4797299 4799098 4801970 4802207 4806105 4807600 4809787 4809793 4810830 4819226 4824129 4826944 4842604 4847052 4847479 4847938 4865270 4865630 4866684 4876079 4881105 4881109 4881122 4881127 4881132 4881136 4881138 4881139 4888526 4889104 4889967 4903457 4907020 4929829 4937092 4956272 4967518 4969758 4971693 4972674 4975612 4986468 4988102 4989784 4989809 4995551 5007650 5007664 5007911 5014319 5016154 5022022 5024472 5026029 5026182 5031134 5036067 5049931 5056104 5056930 5061519 5075581 5082165 5084418 5094779 6180281 6194894 6199412 6199413 6199414 6199416 6199449 6199453 6202496 6203390 6207987 6214522 6232208 6233475 6242752 6246702 6246708 6246969 6251454 6253745 6255677 6257852 6263359 6271102 6274609 Changes incorporated in this version: 4797299 4819226 4824129 4967518 6242752 6246969 6255677 6257852 6271102 6274609 Patches accumulated and obsoleted by this patch: Patches which conflict with this patch: Patches required with this patch: Obsoleted by: Files included with this patch: # upgrade script # Sun customer SUN413B-3310.bin # RAID controller firmware rev 413B patchinfo quick_upgrade.pdf # Flowchart procedures for the experienced saftefw_3310_1170.bin # SAF-TE firmware rev 1170 script_troubleshooting.pdf # Troubleshooting information for the README.113722-11 NOTE: Topics Included in this file: * Problem Description * Upgrading Controller Firmware (4.xx to 4.xx) * Patch Installation Instructions (3.xx to 4.xx) * Prerequisites for Upgrading Firmware * To Install or Upgrade Software on Solaris Operating Systems * To Invoke Sun StorEdge CLI * SAF-TE Firmware Upgrade * Preparing to Upgrade Controller Firmware (3.xx to 4.xx) * s3kdlres Upgrade Script for 3.xx to 4.xx Only * Upgrading the Controller Firmware with a Script (3.xx to 4.xx) * Upgrading the Controller Firmware Manually (3.xx to 4.xx) * PLD Firmware Upgrade * After Upgrading * Special Install Instructions ==================== Problem Description: -------------------- This patch contains firmware which is installed on the StorEdge 3310 RAID array controller and expansion chassis using Sun StorEdge CLI, rather than using installpatch or patchadd. Two types of firmware are included: RAID controller firmware which is installed in RAID controller modules, and SAF-TE firmware which is installed in both RAID controller chassis and disk expansion chassis. RAID firmware 413B addresses the following issues: 4797299 Need data checking facility 4819226 Should have percentage complete for adding drives to an LD 4824129 Request for enhancement of messages on controller FW 4967518 Firmware doesn't track drive failure and the drive can scan back ok. 6242752 Cabling in figure 4.5 and 4.6 does not match table 4.2 6246969 4.11 release has lower performance numbers than previous releases for some of raid1 test cases 6255677 the media scan does not start back up on the rebuilt drive 6257852 LD_RESCUE support removed from SE3000 family without notification 6271102 Secondary RAID controller with disconnected battery reports status as "OK", Write-thru not sustained 6274609 SE3510 with Firmware 4.11I has dirty cache that cannot be destaged to disk (from 113722-10) RAID firmware 411K addresses the following issues: 4702532 write cache not automatically disabled when battery is off line 4714227 The character interface sometimes fails to respond correctly 4714811 Partition/Offset/Size screen should have more information 4722548 Selectable menu items in the firmware have no effect or are N/A to the product 4726561 Aborting Parity Check results in "Parity Regeneration Failed" message 4739050 Lun initialization lost if power failure 4743721 SSCS: no method for identifying disk drives 4799098 NVRAM Restore does not prompt for reboot 4801970 SSCS shows a disk incorrectly under "copy and replace" from FW 4802207 SCSI Bus Reset and daemon.error under heavy load test for 3310 4806105 FW 3.25O enhancement to allow multiple simultaneous rebuilds 4807600 The stripe size on SSCS doesn't show actual information on LDs 4847479 sccli> show inquiry-data does not show correct Firmware Revision 4847938 sccli inquiry does not show the IP address 4865270 Battery status should not be affected by I/O 4866684 open tcp/ip ports 4881105 Ability to set Date, Time, and TZ through the External Interface 4881122 Cached blocks from a deleted logical drive are returned on READ 4881127 Duplicate logical drive IDs assigned on same controller 4881136 Periodic cache flushing feature with user-selectable flush interval 4881138 Write-back caching option with accelerated posting of writes 4881139 Logical drive expansion for RAID 1 logical drives 4888526 Temperature events should contain the sensor number 4889967 Time Stamp reported for an event log entry is not constant, it changes over time 4903457 Deassert Reset on Failed Controller (reinit failed ctrlr) loses access to disks 4907020 Request Sense couldn't get sense data on SE3310 when VEA rescan performed 4937092 The SCCLI/GUI does'nt allow us to set scsi-io-timeout > 10 4971693 controller reset request popup not happeneing when changing scsi timeout value 4972674 inability to rebuild RAID drive if online drive has bad blocks 4976389 Download of SAF-TE fw via SSCS interface says it succeeded even though it didn't 4988102 Port scanning/security software cause the SE3310 to hang 4995551 Serialnumbers in eventlog are inconsistent or missing 5007650 Controller offlines disk drive with an "Unexpected Sense Received" event 5007664 Controller may not corectly understand certain AENs sent out by the drives 5007911 Running sccli to busy minnows can cause SCSI transport messages 5014319 SCCLI> show fru is displaying inconstistant results with 3310 out-of-band. 5016154 SCCLI: oob connection failed with "network connection setup failed" 5022022 DRAM parity errors/SDRAM ECC errors can cause data corruption 5024472 Running both sccli and SSCS inband can hang V240 server 5026029 Firmware upgrade/downgrade shown as controller failure detected from sccli. 5026182 Firmware upgrade/downgrade displayes incorrect times stamp via sccli show events 5031134 single failed drive that affects the bus, can crash the controller 5049931 Connection to 'IP' closed by foreign host with 3.25W fw 5061519 Bus resets encountered when se3kxtr is run on SE3310 SCSI (Minnow) 5075581 show config -x returns inconsistant results (262 invalid events 30 days) 5082165 Drives that are exceeding the block allocation threshold are not taken offline by the controller 5084418 show shutdown-status output inaccurate for 3310 after shutdown controller cmd. 5094779 sccli SE3310 shutdown logical-drive gets controller access error and fails 6180281 vxa5kchk hangs with SE3310 RAID attached 6190524 Disconnected tagged cmd(s) and timeout on glm-attached storage after poweroff cluster node 6194894 Controller requires 2 resets after restoring defaults on 411 6199449 Minnow (3310/3510) RAID does not report correct LUN geometry via Mode Sense data 6199453 LEDs keep blinking after fw downloaded 6203390 lack of sccli oob functionality follows logical drives between chassis 6207987 256k read operation causes unrecoverable controller error 6232208 Firmware option defaults listed in SE3310 customer documentation are out-of-date 6233475 wrong package name in the book "Sun StorEdge 3000 Family 2.0 Software Installation Guide" 6246702 lower cache flush settings required for 3310 4.11 firmware 6246708 3310 Raid controller not responding to disk spin-up command 6251454 SE3310: inband reset also fails on SE3310 6253745 a drive with consecutive bad blocks should not bring down the whole array 6263359 Sun Cluster loses quorum and panics when one node rebooted The 113722-10 README file addresses: 6202496 Synopsis: Incorrect information in patch README (from 113722-09) SAF-TE firmware 1170 addresses the following issues: 4989784 addr IDs are not showing correctly on dual-bus config using sccli 6199412 live downloading SAF-TE package firmware, I/O time-outs could occur for normal data traffic 6199413 Configuration Sensing - Split Bus and SCSI ID values could be wrong 6199414 The temperature reading of a power supply is not valid when a supply is not installed 6199416 The IOM temperature readings were not being reported in the SAF-TE Status page 6214522 SE3310 SAFTE > 1159 generates false power supply and fan events when sccli commands are issued (from 113722-07) SAF-TE firmware 1167 addresses the following issues: 5005317 SE3310's Disk LED is not lit Amber 5014319 SCCLI> show fru is displaying inconsistent results with 3310 out-of-band. 5056104 Unexpected SCSI Reset on 3120 multi-initiator conf with X6758A 5056930 Warning messages in sccli operations with SE HBA(X1018A) 5075581 show config -x returns inconsistent results (262 invalid events 30 days) (from 113722-06) RAID firmware 325W addresses the following issues: 4929829 Format inquiry returns failed to SE3310 from Ultra320 LVD PCI HBA & Sun Fire V440 on-board 4969758 SE3310 uses wrong MAC address for ICMP echo reply under some circumstances 4986468 Giving up messages on SE3310 RAID array LUNs 4988102 Port scanning/security software cause the SE3310 to hang This README file addresses the following issue: 5036067 firmware patch should warn user not to remove FRU for 10 minutes. (from 113722-05) This file addresses the following issues: 4989809 Incorrect information and typo in README.113722-05 SAF-TE firmware 1159 addresses the following issues: 4865630 sccli hangs and causes I/O timeouts when 'show frus' is executed 4956272 PS LED does not change state above 40C 4975612 SAF-TE firmware upgrade results in gross phase signal errors. (from 113722-03) RAID firmware 325S addresses the following issues: 4889104 Logical drive initialization phase takes too long 4881109 Controller hangs upon encountering medium error for Raid 0 or NRAID 4881132 Disk status is incorrect when rebuilding after RAID 1 multiple drive failures 4876079 SE 3310 RAID Array syncs at 80 MBytes/sec on Ultra320 scsi ports 4847052 Silent Data Corruption when ATN line is cut 4842604 SE 3310 controller violates SCSI spec for "receive diagnostic" command 4810830 Spelling error in "view and edit Configuration parameters" menu (from 113722-02) RAID firmware 325Q addresses the following issues: 4826944 Copy and Replace function during rebuild causes data corruption 4809793 scsi incomplete retry errors when cache is optimized for Random I/O 4809787 After "defaults restore" fw asks for a controller unique id 4802207 SCSI Bus Reset and daemon.error under heavy load test for 3310 4792748 3310 incomplete read- retrying w VTS 4764506 Enabling secondary RS-232 port causes serial communication failure (from 113722-01) RAID firmware 325O addresses the following issues: 4717055 Running heavy I/O & scsi-resets to VxVM volumes on SE3310 LUNs may cause data loss 4723464 Diskomizer w/scsi resets on Minnow causes aio write/read errors and long boots ============================================ Upgrading Controller Firmware (4.xx to 4.xx) -------------------------------------------- The following procedure is used for a minor upgrade, such as from controller firmware 4.11 to 4.13. 1. Invoke sccli, using the appropriate syntax for the type of connection used to manage the array: If the array to be upgraded is a RAID subsystem attached to an HBA installed in the local host, type the sccli command with no arguments: sccli If more than one array is connected to local HBAs, sccli will display a list from which the appropriate device may be selected. If the array has an IP address configured, it may be managed remotely using a network (or "out-of-band") connection. To specify this mode, type: sccli [-w ] where is the password assigned to the controller, and is the hostname or IP address of the array. Use single quote characters to quote the password if it contains any characters that are interpreted specially by the shell. If no controller password is set, the -w option and password argument should be omitted. If the hostname for a locally attached array is unknown, the "--oob" option may be specified instead of a hostname, and sccli will try to determine the IP address of a locally-attached array automatically. Verify that you are running sccli version 2.1 or later with the "version" command. 2. Verify the Product and Revision of the array. Type the command show inquiry and examine the output. If the "Product" is "StorEdge 3310F D" and "Device Type: Enclosure" appears, the device is a JBOD chassis and has no controller firmware; proceed to the "SAF-TE Firmware Upgrade" section. If the Product is not "StorEdge 3310," this patch does not apply. Select a different device or discontinue installation of this patch. If the firmware version reported is 413B or later, the RAID controller firmware does not need to be upgraded. If it is reported as "3310 411K vX.XX" or 411, continue with step 3. 3. If the array contains only one controller or the "show redundancy" command does not report that the array redundancy mode is Active-Active and its redundancy status is "Enabled" (dual-controller), the controller firmware upgrade will require that the controller be reset, which in turn will cause the controller to stop servicing I/O requests for several minutes. If one of the previous conditions is true, skip to step 4 and follow the instructions for upgrading a single-controller configuration. If you have a dual-controller configuration running firmware 411K or later, and the "show redundancy" command reports that the Redundancy mode is "Active-Active" and that the status is "Enabled", perform a live firmware upgrade by typing the following command at the CLI prompt: download controller-firmware SUN413B-3310.bin The sccli command will display messages indicating that it is downloading the firmware, programming the controller's flash memory, and "engaging" the new firmware. Wait until the "sccli>" prompt appears again, which may take 10 minutes or more, before proceeding. If you see an error message containing the text: "please repeat download command" or "controller firmware download ok, but engage operation is not allowed," the controller was unable to perform a live firmware upgrade. Wait until it is OK to stop I/O to the array, and then follow the instructions in step 4 to do a non-live firmware upgrade. If no error messages were displayed, proceed with step 5. 4. Single-controller configurations or dual-controller configurations which are in degraded status must be reset after downloading new firmware. First, stop all I/O activity and unmount any filesystems or volumes mounted from the array. Then, at the "sccli>" prompt, type: download controller-firmware -r SUN413B-3310.bin You should see output indicating that the firmware is being downloaded and programmed into flash memory, and then a message indicating that the controller is being reset. Wait until the "sccli>" prompt appears again, which may take 10 minutes or more, before proceeding. 5. Verify that the firmware upgrade succeeded by examining the firmware revision again. At the "sccli>" prompt, type show inquiry and verify that the firmware revision is now reported as "413B". =============================================== Patch Installation Instructions (3.xx to 4.xx): ----------------------------------------------- When upgrading from 3.xx to 4.13, you must have a serial connection to restore settings when the controller firmware upgrade completes, and the serial connection must be set to 38400. When upgrading from 3.xx to 4.13 on Solaris, Windows, or Linux, you must use the upgrade script. Other platforms require the "Configuring the Controller Firmware Manually" instructions contained in this README file. For experienced Sun customers who can use the upgrade script and minimum instructions, use the quick_upgrade.pdf file contained in this patch. Refer to the script_troubleshooting.pdf file provided in this patch if error messages occur when running the upgrade script. ------------------------------------- Prerequisites for Upgrading Controller Firmware (3.xx to 4.xx): ------------------------------------- Before installing controller firmware version 4.13, make sure your environment can support the prerequisites for installing the new firmware including: * Version 2.1 software required Version 2.1 of the "sccli" CLI utility included in the SUNWsscs package, which may be installed from the Sun StorEdge Professional Storage Manager CD or downloaded from the Sun Download Center. If you do not have this package installed, or if you have an earlier version, see "To Install or Upgrade Software on Solaris Operating Systems" later in this document. * Upgrade script required (Solaris, Microsoft Windows, Linux) The version 2.1 software provides the upgrade script for software installations on Solaris, Microsoft Windows and Linux. You must use the s3kdlres script to perform the controller firmware upgrade if you have the 2.1 software installed on Solaris, Microsoft Windows, or Linux systems. This greatly reduces the probability of error and saves time. * Manual upgrade procedure required for HP-UX and IBM AIX systems If you have the 2.1 software installed on HP-UX or IBM AIX systems, you must use the procedure provided in "Upgrading Controller Firmware Manually" in this README file. ----- Note: ---- The manual procedure works on all systems but should not be used on systems which have the upgrade script. * The controller firmware script restores the following configuration settings: * controller-name * controller-id * Logical drive controller assignment (Primary/Secondary) * cache-parameters: write-back, write-through, random, sequential * disk-array: normal-verify, rebuild-verify, init-verify, rebuild-priority * drive-parameters: spin-up, disk-access-delay, scsi-io-timeout, queue-depth, polling-interval, enclosure-polling-interval, auto-detect-swap-interval, smart, auto-global-spare * host-parameters: queue-depth, max-luns-per-id, fibre-connection-mode * channels - This option restores the channel type, (disk or host), and target IDs. * maps - This option restores the LUN mappings. All other settings will be set to the 4.13 default settings. * Ethernet connections required for out-of-band firmware upgrades For RAID controller firmware upgrades, an Ethernet connection (or two Ethernet connections for dual-controller configurations) is required for out-of-band firmware upgrades, if the array to be upgraded is not connected to the host on which the upgrades are to be performed. In order to use the Ethernet connection(s), the controller must have an IP address assigned. * A SCSI connection to the array is required when no network connection to the array is available. * You must stop all I/O to the array before beginning the controller firmware upgrade. * Unmount all StorEdge 3310 logical drives/LUNs from any Solaris hosts connected to the array, for the duration of the install procedure. For example, filesystems mounted from the Sun StorEdge 3310 should be unmounted, and databases accessing raw devices on the array should be shut down. * Do not exceed 32 partitions per logical drive. If your configuration has more than 32 partitions per logical drive, you must back up the data on the affected logical drives, reconfigure the partitions per logical drive to less than 32, redistribute the data across the logical drives, and then upgrade the firmware. * You must reconfigure controller and network parameters after upgrading the controller firmware. For example, the controller password is not restored if you perform a manual upgrade or an inband upgrade using the script. The password is set to null when the upgrade completes. The controller password is restored if you perform an out-of-band upgrade using the script. * You must have a serial connection to restore the IP address, netmask, and default gateway settings for your array when the upgrade completes. The serial connection must be set to 38400. Note the IP address, netmask, and default gateway settings for your array before you begin the upgrade. * Using the RS-232 interface, note the filter map and sector/head/cylinder parameters, and record these settings. The RS-232 serial interface allows additional filter map and sector/head/cylinder parameters beyond what sccli allows. After the firmware is upgraded, there is no way to recover this information. For details, see step 6 in "Preparing to Upgrade Controller Firmware." * If you will be using the script to perform the upgrade: * You must currently have controller firmware version 3.25S or later. * If performing an in-band upgrade using the script, the controller password is not restored. The password is set to null when the upgrade completes. Be sure to specify a more secure controller password. * Due to important data integrity improvements in this release, the performance for RAID1 has been impacted. Random write operations are impacted the most -- performance is reduced by 10-20%. * Stop any Sun StorEdge Configuration Service (SSCS) agent that may be running on a host connected to the array controller. The SSCS agent may interfere with the download of SES code and the script upgrade process. To stop the SSCS agent, issue the shell command "/etc/init.d/ssagent stop" as root on any system where the SUNWsscs package is installed. If the SSCS agent is running on a Microsoft Windows system connected to the array, use the "Services" application (Start->Run->"services.msc") to stop the SSCS Agent service. * Stop the StorADE agent software, or any other software that monitors the array, following the procedures described in the documentation for that application. StorADE, or any other application which monitors the status of the enclosure, may also interfere with controller or SES firmware downloads and the controller firmware script upgrade. * Unpack the patch into any convenient directory (for example, /var/tmp). ============================================================ To Install or Upgrade Software on Solaris Operating Systems: ------------------------------------------------------------ For installation or upgrading instructions for other operating systems, refer to "Sun StorEdge 3000 Family Software Installation Guide," which is available from: http://docs.sun.com/app/docs/coll/3310SCSIarray 1. Go to http://www.sun.com/software/download/index.jsp and click the "Storage Management" link under the "Systems Administration" heading. 2. Click the link for "StorEdge 3000 Family Storage Products--Related Software". 3. Log in or register as necessary. 4. Download and install the appropriate software files. If you are upgrading, uninstall all earlier versions of Sun StorEdge Configuration Service, Sun StorEdge Diagnostic Reporter, and Sun StorEdge CLI, including the agents. 5. To uninstall earlier versions of the software, type: # pkgrm filename where filename for version 2.x is SUNWsscs and filename for version 1.x is: SUNWscsd Configuration Service Agent SUNWscsu Configuration Service Console SUNWscsa Diagnostic Reporter Agent SUNWscui Diagnostic Reporter Config Tool SUNWsccli CLI 6. To install the version 2.1 software package, type: # pkgadd -d . SUNWsscs The Solaris installation package, SUNWsscs, includes the following components: Sun StorEdge Configuration Service agent Sun StorEdge Configuration Service console Sun StorEdge Diagnostic Reporter agent (daemon) Sun StorEdge Diagnostic Reporter Config Tool (UI) Sun StorEdge CLI 7. Provide appropriate responses to each of the installation prompts. 8. After the package is installed, a message is displayed, indicating it was installed successfully. The Sun StorEdge Configuration Service components are installed in the following directories: /opt/SUNWsscs/ssagent /opt/SUNWsscs/sscsconsole The Sun StorEdge Diagnostic Reporter components are installed in the following directories: /opt/SUNWsscs/ssdiagreporterd /opt/SUNWsscs/ssdiagreporterui The CLI is installed in /opt/SUNWsscs/sbin/sccli. 9. Verify that the correct CLI version is installed by typing: /opt/SUNWsscs/sbin/sccli version =========================== To Invoke Sun StorEdge CLI: --------------------------- The firmware procedures require the user to invoke Sun StorEdge CLI (sccli) using the appropriate syntax for the type of connection used to manage the array. If the array to be upgraded is a RAID subsystem, type the sccli command with no arguments: sccli If more than one array is connected to the host, sccli will display a list from which the appropriate device may be selected. If the array has an IP address configured, it may be managed remotely using a network (or "out-of-band") connection. To specify this mode, type: sccli [-w ] where is the password assigned to the controller, and is the hostname or IP address of the array. Use single quote characters to quote the password if it contains any characters that are interpreted specially by the shell. If no controller password is set, the -w option and password argument should be omitted. If the hostname for a locally attached array is unknown, the "--oob" option may be specified instead of a hostname, and sccli will try to determine the IP address of a locally-attached array automatically. ======================== SAF-TE Firmware Upgrade: ------------------------ If you are using SAF-TE firmware revision 1170, then no SAF-TE firmware update is necessary and you should proceed to "Preparing to Upgrade Firmware." 1. Invoke sccli and verify that you are running sccli version 2.1 with the "version" command. 2. At the "sccli>" prompt, verify the SAF-TE firmware version by issuing the "show safte" command: sccli> show safte Ch Id Chassis Vendor Product ID Rev Package --------------------------------------------------------- 0 14 000187 SUN StorEdge 3310 A A000 A000 2 15 005632 SUN StorEdge 3310 D 1170 1170 The "show safte" command produces one line of output for each enclosure; in the example above, a RAID chassis with serial number (000187) has an expansion chassis (005632) attached, but for a standalone JBOD chassis or RAID subsystems with no expansion chassis attached, only one enclosure is displayed. Examine the SAF-TE firmware revision codes displayed in the "Rev" column, and SAF-TE "Package" revision codes displayed in the "Package" column. In the example above, the first chassis indicates revision A000 and the second indicates revision 1170. In this case, only the first chassis is out of date. If BOTH the Rev column and the Package column contain "1170" or higher for every device reported, AND you are sure that no power supply FRUs have been replaced in any of the enclosures listed, no SAF-TE firmware upgrade is necessary; skip to "Preparing to Upgrade Firmware." Otherwise, continue with step 3. 3. At the "sccli>" prompt, type the following command: download safte-firmware saftefw_3310_1170.bin The sccli command prompts you for confirmation and then proceeds with a "live" update of the SAF-TE firmware. It is not necessary to reset the StorEdge 3310 RAID array following the SAF-TE firmware update, but the telnet and serial interfaces on the StorEdge 3310 RAID array may continue to display the old SAF-TE firmware revision until the next controller reset (sccli will display the correct version number). The SAF-TE firmware download procedure takes several minutes to complete. If the selected device is a RAID subsystem with one or more expansion chassis attached, all the attached expansion chassis are upgraded at the same time. If the firmware download operation reports a failure, repeat step 3 one time, after verifying that no other storage software (such as the SSCS agent, StorADE, ESM, etc) is running and possibly interfering with the download. 4. Run the sccli "show safte" command again to verify that the SAF-TE "Rev" column contains version "1170" and the "Package" column reports "1170". For example, sccli> show safte Ch Id Chassis Vendor Product ID Rev Package --------------------------------------------------------- 0 14 012345 SUN StorEdge 3310 A 1170 1170 2 15 005632 SUN StorEdge 3310 D 1170 1170 Continue with step 5 to complete the SAF-TE upgrade procedure. 5. When upgrading SAF-TE firmware earlier than revision 1159, it is necessary to repeat the firmware download command twice, even if it looked like the download operation succeeded the first time. This additional step is required because of a bug in the old firmware that may cause a portion of the new firmware not to be loaded correctly, in which case the power supplies may not be able to report their status correctly. Enter the same command again at the "sccli>" prompt: download safte-firmware saftefw_3310_1170.bin After this second "download safte-firmware" command has completed, run the sccli "show safte" command again to verify that the SAF-TE "Rev" column contains version "1170" and the "Package" column reports "1170". For example, sccli> show safte Ch Id Chassis Vendor Product ID Rev Package --------------------------------------------------------- 0 14 012345 SUN StorEdge 3310 A 1170 1170 2 15 005632 SUN StorEdge 3310 D 1170 1170 6. Repeat steps 2 to 5 for each additional 3310 RAID array subsystem or standalone JBOD chassis. ============================== Preparing to Upgrade Controller Firmware from 3.xx to 4.xx: ------------------------------ If you are using controller firmware version 4.13C, then no firmware upgrade is necessary and you should proceed to "After Upgrading." 1. Back up all data. ------- Caution: ------- To ensure that the existing array is restored using the currently configured settings, back up all data and configuration parameters so you can reference them after the upgrade. The script upgrade procedure does not restore the following settings. * Network settings - IP address, net mask, and gateway settings * Baud rate for the serial port * Sector/head/cylinder settings - See step 6 in "Preparing to Upgrade Controller Firmware" for your recorded parameters. * Certain customized settings - Use the configuration file created in step 5 to verify that all parameters have been restored. See step 5 for details on saving configuration parameters. 2. Change your working directory, if necessary, to the directory in which the patch was unpacked and confirm that the SUN413B-3310.bin file is present using the "ls" command. 3. Invoke sccli and verify that you are running sccli version 2.1 with the "version" command. 4. Verify the Product and Revision of the array. Type the command show inquiry Verify that the "Product" field begins with "StorEdge 3310"; if it does not, this patch does not apply to the selected device. Specify an appropriate device or discontinue installation of this patch. If the "Product" field reports "StorEdge 3310 D" and the "Device Type" field is "Enclosure", the selected device is a JBOD enclosure; proceed to "After Upgrading." If the firmware revision is reported as "0413", check the first 4 digits of the "NVRAM Defaults" string; if it starts with a 4-digit code "413B", the controller firmware does not need to be updated; proceed to "After Upgrading." Otherwise, continue the RAID controller firmware upgrade with the following step. 5. Save the array configuration settings in .txt and .xml formats. These files can be referred to later in case it is necessary to restore or troubleshoot the configuration settings. ----- Note: ----- Configuration setting output files are not transferable. You cannot save a configuration for a specific device and use it to perform the script upgrade on another device. At the "sccli>" command prompt, type: show configuration filename.txt and then type: show configuration --xml filename.xml These commands may take several minutes to complete. 6. Use the RS-232 interface, to view sector/head/cylinder parameters, and record these settings. After nvram is reset, there is no way to recover this information. Record sector/head/cylinder settings by checking the RS-232 connection and viewing "view and edit Configuration parameters > Host-side Parameters > Host Cylinder/Head/Sector Mapping Configuration." ----- Note: ----- The 4.13 firmware automatically allows LDs over 253 GB to be seen by Solaris. After upgrading to 4.13 the sector/head/cylinder values will be set to default values and Solaris will see the proper LD size. However, if you have modified the sector/head/cylinder parameters in 3.2x firmware, after upgrading to 4.13 firmware you should change those parameters back to the values you had set in 3.2x; this is in case a relabel on Solaris is performed at a later date. 7. Check that the array is healthy by running the CLI "show redundancy" command. The dual-controller array is healthy if the Redundancy mode is "Active-Active" and the status is "Enabled." The single-controller array is healthy if the Redundancy mode is "Active-Active" and the status is "Scanning." ================ s3kdlres Script for 3.xx to 4.xx Upgrades ---------------- This s3kdlres script automates the upgrade process from the 3XX code base to the 4XX code base. This script saves the controller configuration in XML format, upgrades the firmware, resets nvram, and restores the configuration with sccli commands based on the XML file. -------- CAUTION ------- DO NOT use the upgrade script if you are performing a minor upgrade, such as 4.11 to 4.13. It must be used only for 3.xx to 4.xx upgrades. The s3kdlres upgrade script is provided in the SSCS/CLI 2.1 software package on the Sun Download Center or on your Sun StorEdge 3000 Family Professional Manager CD. ---- NOTE ---- It is not possible to recover all configuration information with sccli and the XML configuration file. See the LIMITATIONS section in the s3kdlres man page for details. Before running this script it is recommended that you use sccli to save nvram to the local file system, and save the output from "show configuration" for later reference if necessary. To reference the s3kdlres man page on Solaris systems type: man -M /opt/SUNWsscs/man s3kdlres To reference the s3kdlres man page on Linux systems type: man 1m -M /opt/SUNWsscs/man s3kdlres ================================== Upgrading the Controller Firmware with a Script (3.xx to 4.xx): ---------------------------------- The upgrade script works with Solaris, Microsoft Windows, and Linux operating systems only. You must use the upgrade script to upgrade from 3.xx to 4.13 when the 2.1 software is installed on these systems. ------- Caution: ------- Review all procedures and release notes for this major upgrade prior to upgrading your array. Downgrading back to the 3.25 controller firmware is recommended only for a FRU replacement module. All other downgrades risk loss of all data. ----- Note: ----- If you will be upgrading more than one array, take note of the time it takes to upgrade the first array. Then, you can allot the appropriate amount of time it will take to upgrade your other arrays. ----- Note: ----- IBM AIX and HP-UX operating systems must use the manual upgrade procedure. ----- Note: ----- Using the script to perform the upgrade saves time. For example, a simple upgrade was completed by an experienced user in 25 minutes using the script. Using the manual procedure took the same user almost 60 minutes to upgrade the same data. ------- Caution: ------- Before you begin the upgrade, read the "Prerequisites for Upgrading Firmware" section. 1. If you have not already done so, do the following: a. Stop all I/O to the array. b. Unmount all hosts connected to the array. c. Stop any Sun StorEdge Configuration Service (SSCS) agent that may be running on a host connected to the array controller. d. Stop the StorADE agent software, or any other software that monitors the array, Reference the "Prerequisites for Upgrading Firmware" section for more information on the preceding items. 2. For the controller firmware upgrade, you must use the serial port interface to restore the IP address and network settings and to monitor the progress of the upgrade. Use a null modem serial cable to connect the COM port of the RAID array to the serial port on a host workstation. Set the serial port parameters on the workstation to 38,400 baud, 8 bit, 1 stop bit, no parity. 3. Prepare to monitor the progress of the installation by using the serial connection. ----- Note: ----- If an upgrade fails resulting in an indeterminate or incomplete status, contact an authorized Sun Service person immediately. Or, reference the script_troubleshooting.pdf file provided in this patch. 4. To use the s3kdlres script to update the controller firmware, perform the following steps. a. (Solaris and Linux only) To run the s3kdlres script to update the controller firmware, change your directory to /opt/SUNWsscs/sbin and type: ./s3kdlres --device= --password= XML filename is the name of the file where the array configuration output will be saved. Device or IP is the device name or IP address of the array. Password is the controller password, if there is one. If no password is set, do not specify the password option. The following example runs the script and updates the firmware on the specified device, which does not have a password assigned: # ./s3kdlres /var/tmp/xml_conf.out --device=206.111.111.111 The executable script is located in: /opt/SUNWsscs/sbin For detailed instructions about the script, see the "s3kdlres Script Man Page" section below. b. (Microsoft Windows only) To run the s3kdlres script to update the controller firmware, choose Start > Run and type cmd to launch a command-line prompt. Change to the directory containing the s3kdlres script: C:\Program Files\Sun\sccli If you installed the software in a location other than the default location listed above, specify that location rather than what is listed above. And then type: s3kdlres --device= --password= XML filename is the name of the file where the array configuration output will be saved. Device or IP is the device name or IP address of the array. Password is the controller password, if there is one. If no password is set, do not specify the password option. The following example runs the script and updates the firmware on the specified device, which does not have a password assigned: # s3kdlres xml_conf.out --device=206.111.111.111 For detailed instructions about the script on Microsoft Windows systems, double-click the help file, s3kdlres.txt, located in: C:\Program Files\Sun\sccli c. When the script prompts you for the firmware file, specify the path where the firmware patch was extracted and the firmware file name, SUN413B-3310.bin. For example, in Solaris, if you extract the firmware file to the default location, type: /var/tmp/SUN413B-3310.bin For example, in Microsoft Windows, if you extract the firmware patch to C:\Program Files\Sun\sccli, type: 113722-11\SUN413B-3310.bin d. If you encounter an error or exception, refer to the script_troubleshooting.pdf provided in this patch. ----- Note: ----- If you have logical drives on the secondary controller on channels 4 or 5, the upgrade script stops before the LUN mapping is restored, and an error is displayed. No LUNs will be remapped. You must change the logical drive assignments from primary to secondary controller through the RS-232 firmware connection. For details, reference the "Troubleshooting the Script Upgrade" section later in this document. e. When prompted, restore the IP address, netmask, and gateway using the serial interface. If you are using in-band communication, restore LUN mapping for the same device used during the firmware upgrade. f. When prompted, type yes to complete the installation. 5. Verify that the firmware upgrade succeeded by examining the firmware revision again. At the "sccli>" prompt, type: show inquiry and verify that the firmware revision is now reported as "413B". 6. Reference the 3.25 configuration .xml or .txt file saved in step 5 of "Preparing to Upgrade Controller Firmware" to restore any settings that were not reset. The upgrade does not restore: * Network settings - IP address, net mask, and gateway settings * Baud rate for the serial port * Sector/head/cylinder settings - See step 6 in "Preparing to Upgrade Controller Firmware" for your recorded parameters. * Certain customized settings - Use the configuration file created in step 5 to verify that all settings have been restored. Change any parameters that were not restored. 7. To restore your previous LUN host mappings, reference the 3.27 configuration file and run the following CLI commands: a. Restore the logical drive controller assignment using the following command: set logical-drive assignment [primary | secondary] b. Restore the channel IDs using the "configure channel" command. c. Restore LUN maps using the "map partition" command. d. If you use random optimization, restore it with the "set cache-parameter" command. e. Refer to the sccli man page for more information about these commands. 8. Use the RS-232 interface to restore the Host Cylinder/Head/Sector Mapping Configuration, based on the settings you recorded in step 6 in the "Preparing to Upgrade Controller Firmware" section. ========================================================== Upgrading the Controller Firmware Manually (3.xx to 4.xx): -------------------------------------------------------- To use the sccli to do a manual update, perform the following steps. IBM AIX and HP-UX operating systems must use the manual upgrade procedure to upgrade from 3xx to 4.xx. ------- Caution: ------- Review all procedures and release notes for this major upgrade prior to upgrading your array. Downgrading back to the 3.27 controller firmware is recommended only for a FRU replacement module. All other downgrades risk loss of all data. ----- Note: ----- If you will be upgrading more than one array, take note of the time it takes to upgrade the first array. Then, you can allot the appropriate amount of time it will take to upgrade your other arrays. ------- Caution: ------- Before you begin the upgrade, read the "Prerequisites for Upgrading Firmware" section. 1. If you have not already done so, do the following: a. Stop all I/O to the array. b. Unmount all hosts connected to the array. c. Stop any Sun StorEdge Configuration Service (SSCS) agent that may be running on a host connected to the array controller. d. Stop the StorADE agent software, or any other software that monitors the array, Reference the "Prerequisites for Upgrading Firmware" section for more information on the preceding items. 2. For the controller firmware upgrade, you must use the serial port interface to restore the IP address and network settings and to monitor the progress of the upgrade. Use a null modem serial cable to connect the COM port of the RAID array to the serial port on a host workstation. Set the serial port parameters on the workstation to 38,400 baud, 8 bit, 1 stop bit, no parity. 3. You can choose to complete the download in one step. Or, you can complete the upgrade later. You must reset nvram to complete the upgrade. -------- Caution: -------- The download controller-firmware command does not support a live upgrade (a hot download operation). The -r option in the following step must be used to prevent firmware download failure. To complete the download in one step and reset nvram, type: download controller-firmware --reset-nvram -r SUN413B-3310.bin To complete the download without resetting nvram, type: download controller-firmware --dont-reset-nvram -r SUN413B-3310.bin The sccli command will display messages indicating that it is downloading the firmware, programming the controller's flash memory, and "engaging" the new firmware. Wait until the "sccli>" prompt appears again, which may take 10 minutes or more, before proceeding. 4. If you specified the "--dont-reset-nvram" option, run the following command: reset nvram ----- Note: ----- With firmware version 4.13, in a redundant-controller configuration, you can run the reset nvram command on both controllers simultaneously. 5. To re-establish communication with the array, use the serial connection to restore the IP address, netmask, and gateway. Or, remap a partition to LUN0. If you are using in-band communication, restore channel settings (ID), LUN mapping, and inter-controller links for the same device used during the firmware upgrade. 6. Verify that the firmware upgrade succeeded by examining the firmware revision again. At the "sccli>" prompt, type: show inquiry and verify that the firmware revision is now reported as "413B". 7. Reconfigure the array with the RS-232 firmware connection or sccli. ----- Note: ----- The manual upgrade procedure requires extensive configuration because all 4.13 factory defaults have been restored. Reference the 3.25 configuration .xml or .txt file or the RS-232 settings recorded in steps 5 and 6 of "Preparing to Upgrade Controller Firmware" to reconfigure parameters. To configure your previous LUN host mappings, run the following CLI commands: a. Restore the logical drive controller assignment using the following command: set logical-drive assignment [primary | secondary] b. Restore the channel IDs using the "configure channel" command. c. Restore LUN maps using the "map partition" command. d. If you use random optimization, restore it with the "set cache-parameter" command. Refer to the sccli man page for more information about these commands. 8. Use the RS-232 interface to restore the Host Cylinder/Head/Sector Mapping Configuration, based on the settings you recorded in step 6 in the "Preparing to Upgrade Controller Firmware" section. ================ After Upgrading: ---------------- 1. If the SSCS agent was stopped, restart it: on Solaris or other Unix systems, use the command "/etc/init.d/ssagent start"; on Microsoft Windows, use the "Services" utility (Start->Run->"services.msc"). 2. If other software such as StorADE was stopped, restart it following the procedures in the documentation for that software. 3. Whenever a controller is reset, continuous media scans begin on all physical drives that are active components of logical drives, regardless of their media scan status prior to the controller reset. If you want to stop the media scan, using the firmware application, perform the following steps. 1. In the firmware application, choose 'view and edit Logical drives.' 2. Select the logical drive that is currently being scanned. 3. Choose 'Media Scan -> Abort Media Scan,' and then choose Yes to terminate the media scan. 4. Perform steps 1-3 for each logical drive on which you want to stop media scan. Or, to stop the media scan using the Sun StorEdge CLI, run: abort media-check [ ] 4. Use the RS-232 serial interface to confirm that the sector/head/ cyclinder parameters are correct for your configuration. Refer to step 6 in "Preparing to Upgrade Controller Firmware." 5. If you are using Sun StorEdge Configuration Service and Diagnostic Reporter to manage and monitor the storage, the following additional steps are required. a. You must set passwords for all Sun StorEdge Configuration Service users. ----- Note: ----- User passwords are deleted when Sun StorEdge Configuration Service is uninstalled. If you had a previous configuration, you have to reenter the ssmon, ssadmin, and ssconfig passwords. b. Before starting the Sun StorEdge Configuration Service Console, you must run the following command to specify the web browser to access online help. /opt/SUNWsscs/sscsconsole/config_sscon c. The Sun StorEdge Configuration Service agent and the Sun StorEdge Diagnostic Reporter agent (daemon) have been installed but are not configured to start at boot time by default. To enable the Sun StorEdge Configuration Service agent to start automatically when the system boots, and to start it now, Type: /etc/init.d/ssagent enable start d. To enable the Sun StorEdge Diagnostic Reporter to start automatically when the system boots, and to start it now, type: /etc/init.d/ssdgrptd enable start ============================= Patch Installation Instructions: 0 Special Install Instructions: ----------------------------- None. README -- Last modified date: Friday, August 26, 2005