Potential loss of access when removing volumes from an active XRC session

Notification

Risk classification

HIPER (High Impact and/or Pervasive)

Risk categories

Data Access Loss

Affected Domain

DS8900F below R9.1 SP1 - 89.11.33.0
DS8880 below R8.5 SP8 - 88.58.3.0

Abstract

Removing one or more volumes from an active XRC session may cause both clusters to reboot.

Description

If one or more XRC volumes are removed from an active XRC session, using the XDELPAIR command or GDPS Delpair action, it is possible that sidefile entries associated with the relationship may not drain fully before the volumes are removed. This can cause multiple 0x05E0 MLEs and may cause both clusters to reboot.

This problem does not exist if an XRC session is terminated via XEND command, GDPS End action or script equivalent.

Recommended Action

Mitigation

1. Suspend the volumes that are to be removed.

2. Record the UTC timestamp of the last volume

3. Wait until the UTC consistency timestamp is after the last volume was suspended

4. Remove the selected volumes from the XRC session

Fix

Update to the following code levels:

DS8900:
R9.1 SP2 - 89.12.8.0 or later

DS8880:
R8.5 SP8 - 88.58.3.0 or later

Reference ID

350819

Date first published

05 April 2021

[{"Risk Classification":"HIPER","Line of Business":{"code":"LOB26","label":"Storage"},"Business Unit":{"code":"BU058","label":"IBM Infrastructure w\/TPS"},"Product":{"code":"SSHGBU","label":"IBM DS8900F"},"ARM Category":[],"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"All Version(s)"},{"Risk Classification":"HIPER","Line of Business":{"code":"LOB26","label":"Storage"},"Business Unit":{"code":"BU058","label":"IBM Infrastructure w\/TPS"},"Product":{"code":"ST5GLJ","label":"DS8880"},"ARM Category":[],"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"All Version(s)"}]

Tips