HP StoreVirtual 4000 Release Note page 7

The version of a store operation request was not logged prior to the request causing an upgrade
to fail. Logging the version will aid in future troubleshooting.
Enabled the RPC load balancing policy to be changed to round robin from the default of linear
when upgrading from 9.0 to 9.5. Round robin is better for configurations that require higher
sequential read workloads and are not prone to network saturation issues.
VSA configurations sent false warning alarms for high latency values.
During a heavy load of small I/Os, a store process on multiple storage systems could fail in close
succession, causing unreplicated and two-way replicated volumes to be unavailable while the
store process restarts.
A problem occurred if an initiator terminated an iSCSI session while internal login processing
was in progress.
Changed the performance parameters for midline SAS drives to prevent a storage system (node)
from becoming overloaded when there are higher I/O loads.
The createRemoteSnapshot CLI command did not work when the primary management group
for the remote copy was running an earlier version of SAN/iQ and the remote management
group was running a later version of SAN/iQ.
On rare occasions, the creation of an application-managed snapshot (either manual or scheduled)
would hang while the SAN/iQ operating system connected to the application server, which
caused subsequent scheduled snapshots to hang. Now, if there is a problem with snapshot creation,
the action will fail instead.
Version 9.5—Patch Set 02
Centralized Management Console (CMC)
A repaired storage system could not be added as a site for an existing cluster. A message about
the cluster changing from a multi-site to a standard cluster, which typically appears when deleting
a site and the cluster would contain only one site after the deletion, was displayed.
Attempting to edit a site that included a ghost storage system caused the Edit Site dialog box to
become unresponsive.
When email is not configured for a storage system, the warning icon no longer displays incorrectly
next to the storage system icon. The warning icon correctly displays next to the Email icon under
Events storage system (node) in the Navigation pane.
After adding a Failover Manager (FOM) to a management group, the license registration dialog
box is no longer displayed. License registration is only required when adding storage systems.
Clicking OK on the Edit Volume dialog box, even when making changes unrelated to size, caused
the volume size to increase or decrease automatically.
In a management group with 700 or more servers or volumes, the CMC login no longer causes
slow performance.
For non-English systems that use the comma as the separator instead of decimal point, the Edit
Volume dialog box displayed the Reported Size as 10 times larger than it was.
Enhanced the Online Upgrades feature to:
◦
Display a notifications window in the CMC, which ensures you are notified of any updates
to the system, such as patches, customer advisories, or updated release notes. The Notifications
Fixes 7