MCSG: Rolling upgrade rulesΒΆ

Title:MCSG: Rolling upgrade rules
Author:Douglas O’Leary <dkoleary@olearycomputers.com>
Description:MCSG: Rolling upgrade rules
Date created:06/2007
Date updated:08/2009
Disclaimer:Standard: Use the information that follows at your own risk. If you screw up a system, don’t blame it on me...

One of the advantages to using MC ServiceGuard is the ability to selectively patch or upgrade nodes one by one until all the nodes are upgraded. This process is called Rolling Upgrades. Of course, there are specific rules that must be followed to successfully complete a rolling upgrade:

Paraphrased from class notes and Appendix E of Managing MC/ServiceGuard

  1. Cluster configuration files cannot be updated until all nodes are at the same version of the operating system.
  2. All MCSG commands must be issued from the node with the latest version of the MCSG software
  3. Only two versions of MCSG can be implemented during a rolling upgrade.
  4. Binary configuration files may be incompatible.
  5. Rolling upgrades can only be carried out on configurations that have not been modified since the last time the cluster was started.
  6. MCSG cannot be removed from a node while the cluster is being upgraded.
  7. Any new features of MCSG cannot be utilized until all nodes are running the same version of MCSG.
  8. Keep kernels consistent
  9. Hardware configurations cannot be modified

The steps for doing so are:

  1. Halt the node to be upgraded
  2. If necessary, stop the automatic cluster configuration (AUTOSTART_CMCLD=0)
  3. Upgrade the node.
  4. Apply requisite patches
  5. Turn on automatic cluster configuration if desired (AUTOSTART_CMCLD=1)
  6. Rejoin the node with the cluster.
  7. Repeat above for all other nodes.