8100 Release Notes (XCS 6.250)(5697-1121, January 2012)

Version:

XCS 6.250

Description

This document contains information specific to the release of XCS 6.250 for the

EVA4000/6000/8000 and EVA4100/6100/8100. Much of the information included here is

not documented elsewhere, so it is recommended that you read this information thoroughly before

installing and operating the EVAx000/x100.

NOTE: If you are using asynchronous replication, HP strongly recommends that you upgrade to

XCS 6.200 or later to ensure maximum performance.

Update recommendation:

Recommended

Product models

• EVA4000/6000/8000

• EVA4100/6100/8100

These arrays are referenced as EVAx000/x100 throughout the document.

Fixes

The following issues have been resolved in XCS 6.250:

• Addressed a controller termination (termination code 0x0401031c) that occurred when a

DR group member on the destination array was transferred from one controller to the partner

controller during DR group normalization.

• Fixed an issue that caused the DR tunnels to disappear.

• Fixed a possible SPOF (Single-Port-on-Fibre) error that could occur when SCSI commands

timed out on a Red Hat Linux host.

• Fixed a dual controller termination that could occur if the intersite link between the source and

destination arrays becomes unstable.

• Implemented the system inoperative event.

• Fixed an issue regarding increased I/O latency that could occur on a Windows host with a

QLogic HPAK344A 8 Gb HBA when implementing NPIV (NPORT ID Virtualization).

• Fixed an exception that could occur during DR group configuration changes or site failover

resulting in a controller termination (termination code 0x442011F).

• When determining which controller to send a read/write request to, the controller software

did not check for all LUN cache states on the secondary controller and did not see that the

secondary controller was down. Therefore, the request was sent to the incorrect controller,

causing a recursive controller termination (termination code 0x07110100).

• For Vmware ESX 4.1 hosts connected to the array, an issue was fixed where the current

excessive cache correctable error count was unchanged from the previous count, causing

concern about the validity of the error condition.

• Fixed a Unit Stalled Too Long condition and CSM Reset termination. As part of the boot

sequence, the controller software determines its physical position (top/bottom) in the storage

system configuration to identify which controller should process the SCSI traffic. The error

occurred when the controller software attempted to process the SCSI traffic before the software

had determined the physical position of the controller.

Description 3