HP P6000 Enterprise Virtual Array release notes (XCS 10001000) (5697-1819, March 2012)

• Fixed an issue which causes a controller to experience a stalled state and execute a CSM Reset
termination (termination code 0x4238011f) to clear the stalled state due to the tunnel state
transitions not calling back the log for failed merge writes.
• Fixed an issue where a controller experiences a processor exception termination due to a stack
overflow on remote replication write retries when the SCSI-FCP tunnel is open, but the remote
replication array discovery data is marked invalid.
• Added information to events related to invalid group cache operations.
• Fixed an issue where resolving a multiple disk failure condition can leave a DR group unusable.
• Fixed out of sequence replication requests in HP P6000 Continuous Access configurations.
• Added instrumentation to HP P6000 Continuous Access to track instances when a controller
termination (termination code 0x0442011f) occurs due to inter-controller cache inconsistency.
• Fixed an issue where HP P6000 Continuous Access recognizes a host as a possible replication
target, resulting in the host port WWN being deleted.
• Fixed an unexpected resynchronization that could occur during HP P6000 Continuous Access
log removal or invalidation.
• Improved array response to host I/O during HP P6000 Continuous Access copies.
• Fixed an issue where DR tunnel closures caused by SCSI exchange timeouts can lead to Unit
Stalled Too Long or CSM Reset termination in an effort to clear the stalled state.
• Fixed an issue where the SCSI-FCP data replication protocol aborts exchanges if they do not
complete within a specified timeout period, resulting in a Unit Stalled Too Long controller
termination.
• Fixed a problem that results in a controller termination during controller restart if a RSCN (Remote
State Change Notification) ELS (Extended Link Services) request is received prior to the start of
an HP P6000 Continuous Access discovery.
• Fixed an issue where mirror cache is not synchronized when failover clears cache data lost on a
DR group, leading to a termination when proxy I/Os arrive.
• Fixed an issue where the data replication protocol could inadvertently change to SCSI-FCP after
a controller software upgrade.
• Fixed an issue where a controller on the DRM remote array can experience a Unit Stalled
Too Long or CSM Reset termination during a secondary to primary controller LUN transistion.
• Fixed a problem where dynamic expansion of DR logs could create map inconsistency in metadata,
resulting in a recursive controller termination.
Operating system/solution support
This section describes issues affecting array support of operating systems or solutions.
• Fixed an issue observed on Windows hosts configured with QLogic HPAK344A 8 Gb HBAs
where the queue depth counts were being cleared with I/Os active on a connection.
• Fixed events and controller resynchronizations seen with some Brocade products.
• Fixed a problem with abort task and LUN reset responses not being returned correctly to HP-UX
hosts.
• Fixed slow link-up problem for an HP-UX host with an 8 Gb Fibre Channel adapter that is connected
to an EVA4400.
• Fixed issues which prevented proper interaction with Windows 2008 KMS server activation.
Fixes (XCS 10000000) 9