cancel
Showing results for 
Search instead for 
Did you mean: 
Highlighted

SFHA Solutions 6.1: Performing a rolling upgrade of Storage Foundation and High Availability (SFHA)

A rolling upgrade is a method of upgrading the clusters in a system. A rolling upgrade has two main phases. In Phase 1, the installer upgrades the kernel packages and in Phase 2, the installer upgrades Symantec Cluster Server (VCS) and VCS agent packages. Since the process rolls over in phases, it is termed as a rolling upgrade. You can perform a rolling upgrade to minimize the downtime for the cluster. The advantage of this method is that it ensures that the service groups that are online, remain online throughout the upgrade process. You can plan the failover and selectively upgrade systems. This ensures that the failover service groups are never offline.

The rolling upgrade process is as follows:

  1. The cluster is split into two subclusters.
  2. The failover service groups that are online on the first subcluster are failed over to the second subcluster. The parallel service groups on the first subcluster are taken offline.
  3. Phase 1 of the rolling upgrade starts on the first subcluster. The kernel packages are upgraded. Almost all of the packages are upgraded during Phase 1, except for VRTSvcs, VRTSvcsag, VRTScavf, VRTSvcsea and VRTSvbs.
  4. Once Phase 1 of the rolling upgrade is complete on the first subcluster, Phase 1 of the rolling upgrade is performed on the second subcluster. The applications are failed over to the first subcluster. The parallel service groups are brought online on the first subcluster and taken offline on the second subcluster.
  5. During Phase 2 of the rolling upgrade, all remaining packages are upgraded on all nodes of the cluster simultaneously. VCS and VCS agent packages are upgraded. Applications stay online during Phase 2. The High Availability daemon (HAD) stops and starts again.

During Phase 1, the only downtime is the time required for the service groups to failover. The downtime is limited to the applications that are failed over and not the entire cluster.

During Phase 2, downtime occurs when the VCS engine HAD restarts, but this does not include application downtime.

Limitations:

  • A rolling upgrade requires the previous version and target version products to work together. You can perform a rolling upgrade from 5.1 and later versions.
  • If fencing is configured in the environment and an upgrade is done from 5.1SP1RP3 to 6.0 and above versions, the fencing protocol is not compatible between 5.1SP1RP3 and the target.  A patch is required to correct this issue. The patch info can be found on sort: https://sort.symantec.com/patch/detail/7157
  • A rolling upgrade is not compatible with a phased upgrade. You should not attempt to combine a rolling upgrade with a phased upgrade.

Symantec provides both a script-based and a web-based installer to perform a rolling upgrade of Storage Foundation and High Availability (SFHA).

For more information about the rolling upgrade process, see:

Symantec Storage Foundation and High Availability documentation for other releases and platforms can be found on the SORT website.