| Skip Navigation Links | |
| Exit Print View | |
|
Sun Storage Common Array Manager Software Release Notes, Release 6.9.0 Update |
| Skip Navigation Links | |
| Exit Print View | |
|
|
Sun Storage Common Array Manager Software Release Notes, Release 6.9.0 Update |
1. Sun Storage Common Array Manager Software Release Notes
What You Get with the Software
Licensing Optional Premium Features
Best Practices for Browser Performance
2500 Array Support for VMware ESX 4.1 with Firmware 07.35.55.10 and later
6540, 6140, and FLX380 Arrays Upgrade
Sun Storage F5100 Flash Array Notes
Create Snapshot fails using CLI
Controller Smart battery replacement procedure
Array loses communication during registration when IPV6 is enabled
Ancient I/O rebooting due to cache block size
Battery nears expiration: 06.xx write back cache should not be disabled
Both RAID controllers reboot after 828.5 days--2500/6000 arrays
Reboots occur due to ancient I/O being aborted in large configurations
Volume error—evShowVol output: Initialization: Not Completed
Reregister Auto Service Request for CAM
Auto Service Request advisory note: DTS Listener transfer
Array name remains unchanged after a full array reset
No email notification for critical and above events
Access Configuration: Cascaded J4400 issues
Access Configuration: Sun Storage 6Gb SAS HBA—OS hangs
Access Configuration fails with error: java.util.HashMap cannot be cast to java.util.Properties
Access Configuration: non-zoned SAS2 wide port aggregation FRU Report display—F5100 + 6Gb/sec HBA
Array lock key must be set before configuration files are imported
ComponentInsertEvent not generated for disk drive inserted into J4200
Current Jobs page can take more than five minutes to display
Expander firmware 5.3.73 does not persist a SAS zone group
Replication status incorrect when primary volume fails
Service Advisor: replacing failed SIM with dual path configurations
Volumes not deleting from the Single Page window
Attempt to delete partner replication set on secondary array fails
Incorrect sscs add notification description in CLI guide
2500 Series Release Notes: Simplex vs. Duplex Configurations
Disk Locked description is inaccurate in online help
Firmware upgrade fails with error: FWR_UPGRADE_FAILURE,6 for J4200/J4400 arrays
Firmware upgrade fails on J4400 SATA drives
Firmware upgrade fails when expander and disks are upgraded together
Firmware upgrade not recognized by GUI until agent runs
Sun Blade 6000 firmware upgrade fails with down-level expander
Sun Blade 6250 and 6270—upgrading network expansion modules fails
Installation and Upgrade Issues
Error occurs when upgrading 2510 controller and drive firmware
Installation fails on SUSE 10.3 32-bit and 64-bit hosts
Performance Enhancer for 2500-M2
Enabling Try and Buy license using the CLI
“Require-Stop” not working in SUSE11.1 innserv jexec init.d script
Logging in to CAM on SUSE11 using invalid password
Adding or removing SATA drives requires host boot
Sun Storage J4000 and F5100 Flash arrays—Linux hot plug issues
Solaris with Veritas DMP or other host type
Error—”no execute access to opt/SMgr/agent/ notifysmagent.sh”—occurs during system boot-up
UTM LUNs Controlled by “Solaris Traffic Manager” 6000/2500 arrays
Add License error using the CLI
Support Data does not save using Windows 2008 R2 SP1 Servers
The following sections describe known issues and recommended workarounds:
Reference the hardware release notes for more information about known issues that pertain to your arrays.
Bug 15850154—When using the CLI to create a snapshot, the error “failure to find resources” occurs.
Workaround—Use the GUI to create volume snapshots. See “Configuring Volume Snapshots” in the online help or the Sun Storage Common Array Manager Array Administration Guide.
This Smart battery information is a supplement for the CAM Service Advisor Battery Removal/Replacement procedure.
To determine if a Smart battery is installed, verify by checking the following:
Go to Storage Systems > array name > Troubleshooting > FRUs.
From the FRU Summary page, click Battery.
From the Component Summary page for the battery, look at the serial number in the Unique Identifier column.
If the serial number for the battery starts with an "S,” it's a Smart battery.
Under the Name column, click the battery name.
On the Battery Health Details page, you will see Last Learn Cycle and Next Learn Cycle fields. If the battery is a non-Smart battery, these two fields are not present.
When you replace a Smart battery, you do not need to reset the battery age as described in the Service Advisor procedure.
Bug 15744716 (7096012)—The IPV6 Routable address is not being set correctly during array registration. If the CAM host and array are on different subnets, the address defaults to the local link preventing CAM and the array to communicate.
Workaround—You can enable IPV4 for the array, or configure the CAM host and the array to be on the same subnet.
In CAM, go to Storage Systems > array > Physical Devices > Controllers. On Controller Summary page, click Specify Network Configuration under the IPV6 section. To obtain the IPV6 IP address, Routable IP address, and Router IP address for the array, use the netCfgShow command via the array serial port.
Bug 15754288 (7110592)—Firmware 07.80.51.10 can cause ancient I/O reboots if the cache block size does not match the application I/O size.
Workaround—Ensure the application I/O size can fit into one cache block. If the cache block size is too small for the application I/O size, it will result in a shortage of an internal structure known as a buf_t. By setting the cache block size to match the I/O size, the correct number of buf_t's will be available and the ancient I/O will be avoided.
To set the cache block size, go to the Administration page for the selected array.
Firmware revision 07.80.x.x supports the following cache block sizes:
2500-M2: 4k, 8k, 16k, 32k
6x80: 4k, 8k, 16k, 32k
Bug 15742788 (7092652)—The host type setting for AIX hosts with the CAMBEX DPF failover driver should be set to AIX_FO, not AIX as stated in the Sun StorageTek 6140 Array Release Notes.
Bug 15668206 (6983826)—With the 06.xx array firmware, the write back cache is incorrectly inactivated when the batteries enter the near-expiration period. Normally, it should in-activate when the batteries expire.
Resolution —Upgrade the array firmware to 07.xx. See Firmware.
Bugs 15583341, 15640887 (6872995, 6949589)—Both RAID controllers reboot after 828.5 days of continuous operation. A timer in the firmware (vxWorks) called “vxAbsTicks” is a 32-bit (double word) integer that keeps count in the 0x0000 0000 format. When this timer rolls over from 0xffffffff to 0x00000000 (after approximately 828.5 days), if there is host I/O to volumes, the associated drives fail with a write failure.
Original Resolution—Every 24 hours, firmware spawns a task--cfgMonitorTask--that checks the value of the vxworks kernel timing counter. For controllers with 03.xx-06.60 firmware (6000 series) and 03.xx-6.70 firmware (2500 series): Both controllers reboot if counter is greater than 825 days.
Final Resolution—Every 24 hours, firmware spawns a task--cfgMonitorTask--that checks the value of the vxworks kernel timing counter.
This fix staggers the reboots of the controllers for approximately five days so the only impact is a small performance degradation while the reboot occurs.
For controllers with firmware 07.15.11.12 or later (6000 series) and firmware 07.35.10.10 or later (2500 series): Controller A reboots if counter is greater than 820 days. Controller B reboots if counter is greater than 825 days.
Bug 15715109 (7044185)—Windows and Solaris management hosts report communication loss.
Workaround—Register the JBOD locally or use Solaris management host to manage the Oracle Enterprise Linux 6 proxy.
Bug 15626618 (6931169)—In configurations where vdisks include more than 32 volumes, you might see host I/O errors or internal controller reboots from the controller detecting the I/Os that have not been processed within the timeout period (for example, ancient I/Os).
Workaround—A best practice is to quiesce host I/O when performing vdisk re-configuration. This helps avoid host I/O errors or internal controller reboots that might occur before setup operations complete.
Bug 15656194 (6969328)—25xx/6xxx array has a number of volumes with the following in the evfShowVol output: Initialization: Not Completed in the stateCaptureData.dmp file bundled with the support data.
This bug leads to performance issues with RAID 5 configurations.
Workaround—Contact Support at https://support.oracle.com and reference this Bug number.
CAM 6.9 uses a new transport protocol for phone home communications with Oracle. Before you update to CAM 6.9, you must unregister CAM with Auto Service Request (ASR), then reregister with ASR after CAM 6.9 is installed.
To unregister with ASR, log into CAM and go to General Configuration > Auto Service Request > Unregister.
Install CAM 6.9.
You can register with ASR during initial setup, or later from the General Configuration > Auto Service Request > Auto Service Request (ASR) Setup page.
To complete ASR activation, use the My Oracle Support web site: http://support.oracle.com. For more information about ASR activation, see the CAM online help topic “Activating ASR with My Oracle Support.”
Refer to the new DTS URLs in the CAM online help topic “Client Security” to verify your firewall settings.
When performing a firmware update or using Service Advisor to put an array into service mode, messages similar to these may appear in the CAM event log:
Nov 3, 2011 10:32:34 AM com.sun.storage.cam.service.notif.asr.dts.DTSListener transferAttemptFailed SEVERE: Message transfer attempt failed: HTTP Error: 404 Not found Queue not found Nov 3, 2011 10:32:34 AM com.sun.storage.cam.service.notif.asr.dts.DTSListener transferAttemptFailed SEVERE: Retrying... Time To Live in milliseconds = 960
Nov 3, 2011 10:32:34 AM com.sun.storage.cam.service.notif.asr.dts.DTSListener messageSendFailed SEVERE: Failed to send message.
The CAM operation is not affected and these messages may be ignored. Refer to knowledge article 1381185.1 for more information.
Bug 15774381 (7147538)—After performing a full array reset from the Administration page, the CAM array name is not changed to “unlabeled” as was done in previous releases.
Workaround—To set the array name, go to the Administration page and click Save.
Bug 15750331 (7104329)—Configuring email notification for Critical and Down alarms does not send email notification as expected.
Workaround—Refer to the Alarm Summary page and Event Details page to get information about major events that occur on the array.
Bugs 15621203, 15621748, 15621899 (6924428, 6925163, 6925341)–-Auto-save zone template does not import; cascaded J4400s return incorrect PHY port data; hosts connected to ports with no zones see all disks in J4400.
Workaround—Use only the “HostOrSimLinkIn” port to cascade J4400 arrays.
Bug 15640353 (6948920)—When both 6Gb SAS ports are connected to a single expander, a X8 wide port results. The OS hangs with dual ports in single-zoned domain. This configuration is invalid.
Workaround—Disable X8 wide port configurations in Access Configuration Zoning.
Bug 15624460 (6928490)—This error has been observed in a configuration of cascaded J4200 and J4400 arrays with dual-path to the host (one HBA connected to SIM0 domain and another HBA of the same host connected to SIM1).
Workaround—Retry the operation.
Bug 15649467 (6960746)—Report for configuration with both HBA ports into single SAS Domain with Zoning disabled yields FRU Report issues such as:
Chassis.00 FRU reports: One of the two connected Expander 1 Port 0 Cable Status fields is dropped from the report.
Chassis.00 FRU reports: The second connected port Expander 1 Port 1 Cable Status displayed is Degraded.
Access Configuration pages are missing port 0.
Attaching both ports from an HBA into the same F5100 expander results in a “wide port” (8 phys). Sun Storage Common Array Manager software models the aggregated connections as a single, logical SAS port. The information displayed in the Access Configuration Summary only lists the connected port with the higher alphanumeric sort. For example, if Ports 0 and 2 are connected into the same expander, only Port 2 is listed.
This issue also occurs in the CLI where Port options list a single entry under the Connector list.
Other forms of addressing the wide port are still in force. The SAS Address of the port is available for use in operations, presenting a single address for all eight PHYS of the two HBA connectors.
If the configuration file to be imported contains secure volumes, the array lock key must be set before the import. The verification step of the import fails if a secure volume is found but the array lock key has not been set. The import job will not be started and no settings on the target array will be changed.
Bug 15644125 (6953638)—When a disk drive is inserted into the J4200, no ComponentInsertEvent is generated and entered into the Event Log. Only ValueChangeEvents [such as (Status +) from Removed to OK for disk in slot x] are generated.
Bug 15582215 (6871197)—In a Sun Storage 6180 array with 1023 Volume Copies, it takes over five minutes for the Current Jobs page to display.
Workaround—Click the Current Jobs page a second time to display the page more quickly.
Sun Storage F5100 Flash array expander firmware 5.3.73 does not persist a SAS zone group in the case where an initiator's SAS address changes. Initiator SAS address change can be caused by a host reboot or when a new initiator is plugged into a previously zoned F5100 port. When a SAS address changes, Sun Storage Common Array Manager will still report the initiator and the devices are associated; however, the F5100 expander has actually disassociated the host from the zoned devices. This results in the host losing access to zoned devices.
Workaround –Re-zone the new initiator address with the existing devices, to ensure the new host initiator is associated with the target devices. This issue is fixed in F5100 expander firmware 5.4.4.
Bug 15399510 (6561709)—When the primary volume in a replication set (6xxx arrays) fails, the management software might incorrectly list the volume as replicating.
If you have dual paths to hosts in zones with cascaded JBODs and you need to replace a failed SIM, you might encounter a problem when you have to plug in the SAS cables to the new SIM. This is because the new SIM is not zoned and all hosts see all disks until the old zone is restored.
Workaround—Before you replace a failed SIM, be sure to stop all I/O, even in dual path configurations. When you insert the new SIM, you have to upgrade the firmware which requires stopping the I/O anyway. Taking this step before replacing a failed SIM enables rezoning the new SIM before attempting further I/O.
Bug 15542935 (6807053)—Unable to delete volumes with the “Show Data in Single Page” option from the Volume Summary page. After multiple volumes are deleted, the Volume Summary Page still shows the same number of volumes as before.
Workaround—Use paginated view to delete volumes.
Bug 15723678 (7057616)—When you delete a replication set using the CLI or GUI, the partner replication set on the secondary array will not be deleted if the controller that is participating in the replication has a port that is down.
Workaround—Go to the secondary array and delete the partner replication set from that array.
Bug 15798792 (7177830)—The sscs add notification command on page 32 of the CAM CLI guide (http://docs.oracle.com/cd/E24008_01/pdf/E24015.pdf ) incorrectly states the default sends out all alerts.
Workaround—The default sends an alert type of ‘clear'. Change the text to:
-m,--alarm-level <down|critical|major|minor>Specifies the minimum priority level of alerts to be sent out. By default, an alert of type 'clear' will be sent out.
The Sun StorageTek 2500 Series Array Release Notes, Release 1.4 document contains contradictory information on pages one and 21 concerning the StorageTek 2501 Expansion Module in simplex and duplex configurations. The correct information is this: Only Simplex 2501 expansion modules may be attached to a Simplex 2500 series controller tray.
Bug 15772650 (7145187)—The field description for Disk Details > Disk Locked is inaccurate.
Correction—For disks supporting Data Encryption Services, a value of True means this disk is locked and cannot accept read or write operations. False means the disk is not locked.
For notable fixed issues related to firmware, see Notable Fixes.
Bugs 15582208, 15617124, 15621944 (6871188, 6919285, 6925388)—Upgrading firmware on J4200/J4400 attached to x6250 blade fails, with the following error message:
Error upgrading firmware. Check the following error message and array alarm page for possible problems. flash firmware image failed FWR_UPGRADE_FAILURE,6
The firmware on the JBOD remained at 3R21 and all disks remained at the same FW level.
Workarounds—For this symptom–SUSE Linux host, J4400 SIM upgrade failed with return code 6, possibly leading to SIM firmware mismatch–there are two possible workarounds:
If the upgrade failed without firmware mismatch, do this:
Power cycle the array and rerun the firmware upgrade wizard GUI.
If the upgrade failed with firmware mismatch, do this:
Power cycle the array and rerun the firmware upgrade using one of the following CLI commands:
sscs modify -a <array-name> -f -t sim -o -w firmware
or
csmservice -i -a <array-name> -f -t sim -o -w
Bug 15633465 (693982—When upgrading a J4400, the SATA disk firmware upgrade fails.
Workaround—Move affected disks to another slot before upgrading so they do not remain in a degraded state.
Bug 15614975 (6916355)—Errors occur when you attempt to upgrade expander and disk firmware at the same time.
Workaround—Power-cycle the array and re-run the firmware install wizard. Then resume the disk drive firmware upgrade.
Bug 15583664 (6873568)—After a firmware upgrade, the CAM GUI is not updated with the correct version.
Workaround—Wait 5 minutes for the agent to run or run the agent manually.
Bug 15643397 (6952753)—During a Sun Blade 6000 firmware upgrade, the following error was reported:
No such expander - 50800200006deabf SIGSEGV in Linux libSTORARC.so
Workaround—Use a non-Adaptec HBA (mixed blade environment with SAS controller).
Bug 15639625 (6948014)—An attempt to upgrade Sun Blade 6000 firmware fails if the associated NEM is not at current revision level.
Workaround—When the Network Expansion Module (NEM) and Sun Blade 6000 Disk Module are flagged for upgrade, execute an upgrade on the NEM first. Then, the upgrade works for the Sun Blade 6000 Disk Module storage. For more information, see the SAS Compatibility Firmware Guide.
Bug 15653326 (6965677)—CAM reports that the upgrade of the NEM expanders completed successfully. However, the updated firmware is not reflected in the GUI.
Workaround—Retry the upgrade (multiple attempts might be needed) or use a non-Adaptec HBA.
Bug 15713795, 15771749 (7042337, 7143862)—When you upgrade controller and drive firmware for the StorageTek 2510 array, the upgrade fails and the Install wizard displays “Error upgrading firmware.”
Workaround—Perform the upgrade in two steps. First upgrade only the controller firmware, and then upgrade the drive firmware.
Bug 15755131 (7112028)—Installation of CAM fails with the error:
The package jdk- did not install successfully.
Workaround—Run uninstall -f and then reinstall the CAM software. The uninstall removes the jdk so you do not have to remove it manually.
Bug 15820643—Currently, the Try and Buy or Evaluation option for Performance Enhancer does not allow you to disable the evaluation setting.
Workaround—Either replace the trial license with a permanent license or wait until the evaluation period expires.
Bug 15874509—To enable Try and Buy or evaluation licenses using the CLI, log into the system using the sscs command and issue sscs enable --help license to see the command format.
Enter the command sscs enable -a <arrayName> -e <FeatureName> license', where <FeatureName> is PerformanceEnhancer or ReplicationSet.
Note - Before installing CAM on Oracle Enterprise Linux 6.0 or Red Hat Enterprise Linux 6.0, you must install the packages listed in the “Linux OS Requirements” section of the Sun Storage Common Array Manager Installation and Setup Guide.
Bug 15692963 (7013981)—The dependency boot sequence LSB comment for “Require-Stop” is missing in the /etc/init.d jexec script. Currently, this generates warnings whenever an administrator installs other unrelated products such as RPMS, HBA management applications, or failover RDAC packages.
Workaround—Add the following line in /etc/init.d/jexec:
# Required-Stop: $local_fs
Bug 15611581 (6911829)—If you enter an invalid password when logging into CAM on SUSE11, login ceases to work even when valid password is provided.
Workaround—Close the browser, then open it again and log in to CAM using the correct password.
Bug 15559493 (6835314)—Linux kernel:mptbase(Abort) occurs on both RedHat 5.2 and SUSE 10.
Adding or removing SATA drives to a Linux host yields the following message:
kernel: mptsas: ioc2: removing sata device: fw_channel 0, fw_id 4, phy 4, sas_addr 0x500163600010a390
kernel: mptbase: ioc1: LogInfo(0x31120101): Originator={PL}, Code={Abort}, SubCode(0x0101)
The following message is also common:
multipathd: sdd: checker msg is “tur checker reports path is down”
Normally the (kernal: mptsas) messages are seen for zoning operations but when the (kernel:mptbase) message is also seen, only a reboot will recover the hosts ability to properly see the drives.
Workaround—Reboot host.
Two categories of J4x00/F5100 Hot Plug issues have been identified--those that require a Linux host reboot and those in which the hot plug is not supported. These are separated into two sub-sections:
Linux host reboot is required if the array or any of its components are not recognized after changes such as cable re-configurations, firmware upgrades, etc.
StorageTek SAS RAID HBA hot plug is not supported.
Sun StorageTek SAS RAID HBA is inconsistent with reports and missing FRU disk information. This is expected J4x00 registration behavior during SAS bus re-scan.
Firmware upgrade failed for Sun Blade 6000 Disk Module.
J4x00—Name issues during registration and upgrade connected to B0/B1 side.
J4x00—Linux J4x00 registration—sg_map -i must be clear of all errors or it fails; leads to host reboot to clear sg_map.
Newly added or removed targets (i.e. disks) due to changes in Access Configuration, cable re-configuration, firmware upgrades, or the addition of new storage can potentially cause the host to hang or panic due to known Linux kernel issues.
Bug 15498034 (6731595)—J4200/J4400: Sun StorageTek PCI-Express SAS Host Bus Adapter B3: SUSE 9 SP4: Multipath (device mapper): Alternate SIM failure causes host to hang.
Bug 15498590 (6732411)—J4200/J4400: Sun StorageTek PCI-Express SAS Host Bus Adapter B3: SLES9SP4: Multipath: Lost Communication to JBOD after Sun Storage Common Array Manager software SIM firmware upgrade.
Bug 15525921 (6777089)—MPT 3.16.00.00 pandora hba SAS cable insert, expander reset or power cycle hang x86 host until reset.
Bug 15549335 (6817878)—OS does not see zoning filtering through to proxy server correctly. This issue is specific to Linux hosts.
Bug 15556018 (6830042)—As a result of JBOD expander firmware upgrade, expander reset might hang Linux OS.
Bug 15558198 (6833156)—1.26.03.00: linux missing targets, fdisk -l and sg_map -i hangs after JBOD second power cycle.
Sun StorageTek SAS RAID HBA hot plug is not supported. Generally, the workaround is to reboot the array between any cable changes etc.
Bug 15493007 (6723686)—J4x00 failed disk drive (Sun StorageTek SAS RAID HBA) not reported.
Bug 15498768 (6732704)—J4x00 Windows agent (Sun StorageTek SAS RAID HBA) 0.0 MB reported; drive FRU missing unique identifier.
Bug 15840516—With the release of firmware 07.84.44.10, the host type ‘Solaris (with Veritas DMP or other)' is no longer a valid host type.
Workaround—If you are using Veritas with DMP, refer to Veritas support (http://www.symantec.com/support/contact_techsupp_static.jsp) for a recommended host type. http://www.symantec.com/support/contact_techsupp_static.jsp
Bug 15629617 (6934913)—This error occurs when /opt(SMagent install directory) is built on a disk partition other than root partition.
Workaround—Though event notification is lost as a result, there are no adverse effects because the SMagent detects the disk device(UTM volume) during re-scan. The notifysmagent.sh script should be available to the syseventd since the script is located in the same directory of the SMagent startup script.
After you upgrade to S10U3 (or later), the in-band management UTM LUNs are controlled by Solaris Traffic Manager (MPxIO). In most cases, in-band management will not fail as a result of this; however, it is best practice to ensure that the UTM LUNs are not controlled by MPxIO.
Workaround—Use the format inquire command to get the eight-character Vendor (VID) and Product IDs. Follow this procedure:
Edit the file /kernel/drv/scsi_vhci.conf
The following line should read:
device-type-scsi-options-list = “SUN Universal Xport”, “disable-option”; “disable-option” = 0x7000000
Run the stmsboot -u command.
Respond to the prompts as follows:
WARNING: This operation will require a reboot.
Do you want to continue? [y/n] (default: y) y
The changes will result when the system is rebooted.
Reboot the system now? [y/n] (default: y) y
Bug 15833559—When adding a premium feature license using the CLI, the error ‘not found/read' occurs.
Workaround—Add the license using the GUI.
Bug 15734264 (7076972)—Application support data files cannot be saved using Windows 2008 R2 SP1 servers. Support data for an array is collected and saved in the Common Array Manager by selecting array > General Configuration > Support Data.
Workaround—On the Windows 2008 server, disable “Do not save encrypted pages to disk.” Go to Tools > Internet Options > Advanced > Security and uncheck "Do not save encrypted pages to disk."
|