HP ProLiant Cluster HA/F500 for Enterprise Virtual Array Enhanced DT Supplement Guide

Disaster Recovery
Source Site Failback
The two failover-only options for Vdisk creation allow the host to control when a
Vdisk moves to a preferred path. For example, if path A is preferred and that path
becomes unavailable, path B is used. The host will then control the movement back
to path A when it becomes available later.
EVA Storage Failback Procedure
Problem
EVA storage fails at one site requiring a site failover:
•
•
•
•
•
•
Manual failover to the alternate site is successful.
Cluster is successfully running on the alternate site.
Investigate the source of the EVA storage problem at the original site and correct.
During the boot process of the original storage subsystem, the cluster temporarily
loses access to the disks that had to be failed over.
Cluster loses the reservation of the disk and fails.
Cluster requires a manual restart of the cluster service and manually brings the
drives back online.
Solution
To overcome the disruption of cluster service during the recovery of the storage
subsystem:
1. Disable the ISLs (Intersite Links) between the sites.
Temporarily disable the ISLs between the sites so that host access can
temporarily be removed from the original failed storage subsystem.
2. Restore power to the EVA storage at the failed site.
HP ProLiant Cluster HA/F500 for Enterprise Virtual Array Enhanced DT Supplement Guide 3-5
HP CONFIDENTIAL
Writer: Woody Jernigan File Name: d-ch3 Disaster Recovery.doc
Codename: 49er IV Part Number: 339223-002 Last Saved On: 9/30/03 10:48 AM