When performing shutdown and maintenance on a LifeKeeper-protected server, you must put that system’s resource hierarchies in service on the backup server before performing maintenance. This process stops all activity for shared disks on the system needing maintenance.

The typical maintenance process for a two-node LifeKeeper cluster (consisting of servers Server A and Server B) is described below.

  1. Disable resource monitoring on Server A. Set LKCHECKINTERVAL=0 in /etc/default/LifeKeeper on Server A and execute killall lkcheck. Note the current value of LKCHECKINTERVAL as it will be used in step 8 below.
  2. Bring all resource hierarchies in service on Server A. On the primary server, Server A, use the LifeKeeper GUI to bring in service any resource hierarchies that are currently in service on Server B. This will unmount any file systems currently mounted on Server B that reside on the shared disks under LifeKeeper protection. See Bringing a Resource In Service for instructions.
  3. Stop LifeKeeper on Server B. Use the command $LKROOT/bin/lkcli stop to stop LifeKeeper on Server B. Your resources are now unprotected.
  4. Shut down Linux and power down Server B. Shut down the Linux operating system on Server B, then power off the server.
  5. Perform maintenance on Server B. Perform any necessary maintenance on Server B.
  6. Power on Server B and restart Linux. Power on Server B, then reboot the Linux operating system.
  7. Start LifeKeeper on Server B. Use the command $LKROOT/bin/lkcli start to start LifeKeeper on Server B. Your resources are now protected.
  8. Enable resource monitoring on Server A. Set LKCHECKINTERVAL=<previous value from step 1> in /etc/default/LifeKeeper on Server A and execute killall lkcheck.
  9. Disable resource monitoring on Server B. Set LKCHECKINTERVAL=0 in /etc/default/LifeKeeper on Server B and execute killall lkcheck. Note the current value of LKCHECKINTERVAL as it will be used in step 16 below.
  10. Bring all resource hierarchies in service on Server B. On the standby server, Server B, use the LifeKeeper GUI to bring in service any resource hierarchies that are currently in service on Server A. This will unmount any file systems currently mounted on Server A that reside on the shared disks under LifeKeeper protection. See Bringing a Resource In Service for instructions.
  11. Stop LifeKeeper on Server A. Use the command $LKROOT/bin/lkcli stop to stop LifeKeeper on Server A. Your resources are now unprotected.
  12. Shut down Linux and power down Server A. Shut down the Linux operating system on Server A, then power off the server.
  13. Perform maintenance on Server A. Perform any necessary maintenance on Server A.
  14. Power on Server A and restart Linux. Power on Server A, then reboot the Linux operating system.
  15. Start LifeKeeper on Server A. Use the command $LKROOT/bin/lkcli start to start LifeKeeper on Server A. Your resources are now protected.
  16. Enable resource monitoring on Server B. Set LKCHECKINTERVAL=<previous value from step 9> in /etc/default/LifeKeeper on Server B and execute killall lkcheck.
  17. Bring any desired resource hierarchies back in service on Server A. On the primary server, Server A, use the LifeKeeper GUI to bring in service any resource hierarchies that should be in service on Server A. See Bringing a Resource In Service for instructions.

Feedback

Thanks for your feedback.

Post your comment on this topic.

Post Comment