2009.Q3.2.0

This minor release contains a number of bug fixes for the 2009.Q3 release. This update is recommended for all users, particularly those using NDMP, iSCSI (especially with iSNS), shadow migration, clustering, and Infiniband. Customers with clustered systems should see RN017 at the bottom of this page before upgrading their systems.

Because this minor release includes all features and bug fixes included in previous releases of 2009.Q3 (including 2009.Q3.0.0 and 2009.Q3.1.0), customers updating from older releases should review the release notes for all previous 2009.Q3 releases starting with 2009.Q3.0.0. This is particularly important for customers using iSCSI as there are significant iSCSI management changes in 2009.Q3.

The following major issues have been addressed:

• NFS state (including locks in v3 and v4, and delegations in v4) not persisted across takeover/failback
• Shadow migration with large numbers of threads causing poor responsiveness from the management interface
• Shadow migration using NFSv3 obtaining the wrong permissions for files
• Shadow migration panic with CIFS clients
• Unable to add storage to a raidz2 pool with wide stripes
• Unable to create LUNs on VMware and VirtualBox simulators
• Several stability and functionality issues with iSCSI and iSNS
• Potential security issue with non-root administrative users: in certain error conditions, the appliance shell can give authorized non-root administrative users an OS shell on the system. Such a shell runs with the uid associated with the user, so the user cannot make any changes to the system configuration nor read or write any files that the user is not authorized to read and write according to filesystem permissions. However, this may allow access to files that they wouldn't have been able to access due to other policy constraints, such as filesystems exported only to specific hosts to which the user does not have access.
• NDMP backups sometimes omit files in directories with many subdirectories having long names. This issue is very unlikely to occur but results in silent failure to backup some files, so customers should be sure to verify the contents of full backups created on releases prior to 2009.Q3.2.0 or else replace them with full backups created using this release or later.
• Performance issues related to disabling LSO for Infiniband: previous updates disabled LSO due to data integrity issues, but this had a major impact on Infiniband performance. The integrity issues have been addressed in this release, and LSO has been enabled.

Supported Platforms

• Sun Storage 7110
• Sun Storage 7210
• Sun Storage 7310
• Sun Storage 7410
• Sun Storage 7000 Simulator

Issues Addressed

The following CRs have been fixed in this release:

Known issues

The Configuration/Network view permits a wide variety of networking configuration changes on the Sun Storage system. One such change is taking an existing network interface and associating it with a different network datalink, effectively moving the interface's IP addresses to a different physical link (or links, in the case of an aggregation). In this scenario, the network routes associated with the original interface are automatically deleted, and must be re-added by the administrator to the new interface. In some situations this may imply loss of a path to particular hosts until those routes are restored.

The Configuration/Network view permits a wide variety of networking configuration changes on the Sun Storage system. One such change is modifying or deleting the active administrative interface over which the administrator is connected to the web browser or command-line interface itself, which may also be providing the default route for the system. The appliance software attempts to detect such modifications and warn the administrator first. If such a reconfiguration is confirmed or not detected in advance, the administrator will lose contact with the appliance during the network reconfiguration. The network reconfiguration will complete, but if contact cannot be restored with the appliance (by reloading the web address or reconnecting using ssh) due to an error in the configuration applied by the administrator or the loss of the default route, the administrator should log in using the service processor console and correct the configuration using the command-line.

In the current version of J4400 Array firmware, used in the disk enclosures for the Sun Storage 7410, the J4400 emits an audible alarm when a disk drive is hot-unplugged. The alarm warns the operator that a drive has been removed, and the alarm is silenced when the drive is replaced, either by the original drive or by a new replacement. In some operator scenarios, it may be preferable to not emit such an alarm, particularly if a spare drive is not immediately available. The operator can manually silence the alarm by depressing a button on the J4400 front panel. Future versions of the J4400 firmware and appliance software will disable this behavior and/or permit it to be configured by the administrator.

When reconfiguring the Active Directory service to participate in a different domain, restart the CIFS service before attempting to reconfigure the Active Directory service.

In a 7210 System, removing the first system disk will make the system unbootable, despite the presence of a second mirrored disk. To workaround this issue, break into the BIOS boot menu, under 'HDD boot order', modify the list so the first item is "[SCSI:#0300 ID00 LU]".

With a large number of network interfaces, the CIFS service NetBIOS registration mechanism can time out, causing the CIFS service to fail. If you encounter this situation, please contact Sun support to work around the issue.

Under heavy load in large configurations, the first SIM card (SIM 0) can fail. The symptoms are a blue LED on the card itself and an audible alarm, with possible alerts in the UI regarding paths and/or power supplies being removed from the chassis. I/O will continue down other available paths, and there no impact to availability, though performance may suffer. Re-seating the SIM card (removing it and inserting it) should fix the problem. If this problem persists, please contact Sun Support.

If an automatic snapshot is deleted during replication of the same project, it's possible for the replication to fail. The safest workaround is to disable automatic snapshots if a project is being replicated. Alternatively, picking a schedule that is guaranteed to not interfere with ongoing replication may work, but is dependent on each incremental replication stream being sent in a well-defined period of time.

Microsoft Internet Explorer 8 is currently not supported for the web-based management interface.

NDMP backups of appliance data includes share metadata, including projects, shares, and share properties (see the documentation for details). However, user quotas (new to the 2009.Q3 release) are currently not backed up with NDMP. This configuration will have to be manually restored.

NDMP does not backup nested filesystems (new to 2009.Q3). When creating filesystems mounted within other filesystems, be sure to specify each mountpoint separately in the NDMP DMA backup configuration.

The NFS/RDMA support in 2009.Q3 is a preview for early adopters and should not be used on production systems. There are currently no supported clients. Linux clients still under development have experienced data validation errors and loss of access to NFS/RDMA shares on the appliance.

The 2009.Q3 release introduces target groups and initiator groups on which individual iSCSI LUNs are exported. When replicating projects containing iSCSI LUNs using the built-in remote replication feature, ensure that the target groups and initiator groups used by LUNs within the project also exist on the replication target. It's only required that groups exist with the same name, not that they define the same configuration. Failure to ensure this can result in failure to clone and
export replicated LUNs.

In systems with large numbers of network interfaces using jumbo frames, some network interfaces may fail to come up due to hardware resource limitations. Such network interfaces will be unavailable, but will not be shown as faulted in the BUI or CLI. If this occurs, turn off jumbo frames on some of the network interfaces.

Live cable and/or subnet manager partition re-configuration may result in mis-configured IPoIB network devices, datalinks or interfaces. If this happens, delete the affected IPoIB interfaces and partition datalinks on the affected controller. To recover the underlying IPoIB device, make sure that all IB cables are properly configured and the subnet manager is up and running with a valid partition table. Next, reboot the controller with the mis-configured IPoIB device.

After performing a configuration restore, the default route may become inactive. It may need to be reconfigured.

Clustered systems must be upgraded to the 2009.Q2.5.0 or any 2009.Q3 release before being updated to the 2009.Q3.2.0 release. If a cluster is already running either 2009.Q2.5.0 or later (including any 2009.Q3 release), it may be upgraded directly to 2009.Q3.2.0. Otherwise, it must first be updated to 2009.Q2.5.0. The 2009.Q2.5.0 release can be downloaded from the Software updates page.