HP 3PAR InForm OS 3.1.1 GA/MU1/MU2/MU3 Release Notes

DescriptionItemIssue ID
When an ATS request is initiated while DCOW is
pending, the ATS handling path does not resolve the
pending DCOW and causes data corruption.
Import of a snapshot in ESX fails with invalid
meta data when VAAI ATS is enabled.
69127
In cases where NL drives are overloaded to the extreme,
upper layers can see this as loss of access. If multiple NL
In cases where NL drives are overloaded to
the extreme, upper layers can see this as
69606
drives are in this state, this loss of access can result inloss of access. If multiple NL drives are in
volumes going into preserved state. This is resolved tothis state, this loss of access can result in
volumes going into preserved state. prevent this type of overloading. It also allows longer NL
access times and multiple retries to ensure that a loss of
access is actual and not a false alarm.
Unresponsive Transient MCU and I2C error conditions
do not generate and clear Component State Change alerts
Unresponsive Transient MCU and I2C error
conditions no longer generate and clear
69908
when the underlying condition clears during retry.
Persistent error conditions are not affected.
Component State Change alerts when the
underlying condition clears during retry.
This resolves an issue in Inform OS that could cause node
panic during cluster join operation. This issue exists in
Inform OS versions before and including 3.1.1.MU2.
Resolves an issue that can cause node panic
during cluster join.
69682
If a network message is sent to a node that is removed
from the system, the networking infrastructure becomes
Network issue when node is added or
removed.
70121
unresponsive. Any future network issue cannot be handled,
and it results in connectivity loss to the cluster until the
network interface is restarted.
This issue is resolved. The network interface does not
become unresponsive in this situation but eventually times
out, and allows handling of other networking changes.
Changed the recovery mechanism from resetting the
controller node to resetting the HBA port to resolve HBA
firmware error encountered very rarely.
Change recovery mechanism for firmware
error from node reboot to port reset.
70195
In some circumstances, the logical disk tuning phase of
tunesys can cause an out-of-space issue.
The logical disk tuning phase of tunesys can
cause potential out-of-space issue.
70918
The issue is resolved by compacting the Common
Provisioning Group (CPG) associated with each LD being
tuned.
Avoid potential SYSMGR failures by disabling reclaim
for a volume if the total number of vlmaps exceed a
certain threshold.
Avoid potential SYSMGR failures.71127
When a node is rebooted before the system is fully
started, its status is properly reported in shownode.
Node state properly indicated after early
node reboot.
71684
A VV block can get stuck due to missing events.Rare chance of stuck VV block due to
missing events
72649
This addresses high write service time for volumes on F
class and T class systems when remote copy synchronous
Addresses high write service time for
volumes on F class and T class systems when
72976
mode is enabled. If a system sees similar performanceremote copy synchronous mode is enabled
on these volumes. degradation with remote copy synchronization enabled,
then it can be resolved by manually enabling service time.
For more information, contact HP technical support.
In HP 3PAR OS version 3.1.1, modifying the snapspace
during an active physical copy relationship causes
recursive system manager cores. This is resolved.
Modifying snapspace during an active
physical copy causes system manager cores.
73154
Addresses the IO performance issue in systems where
certain storage classes are configured on some node
Address the IO performance issue in systems
where certain storage classes are configured
on some node pairs.
74054
pairs, but not all. Software now erases inconsistent data
of a storage class as if there are a number of such storage
46 InForm OS 3.1.1 MU3 Release Notes