Fallover and fallback

Edit online

GLVM for PowerHA® SystemMirror® Enterprise Edition handles the automation of fallover from one site to another in response to an outage at a local site, minimizing recovery time for the application. When a site fails, the resource group configuration ensures that the geographically mirrored volume groups are accessible from the remote site.

GLVM for PowerHA SystemMirror Enterprise Edition automates application recovery by managing:

  • The fallover of nodes within a site based on node priority (as identified in the node list for a resource group)
  • The fallover between sites (as specified by the site management policy for a resource group)
  • The fallback of a resource group to a node as configured.
When an application is running on a local (or primary) site and the remote site is part of the cluster:
  • Updates to the application data are made to the RPVs associated with the geographically mirrored volume groups
  • Data is mirrored using the GLVM on the RPVs.
If the node or the disks at the local site become unavailable:
  • The application moves to a server at the remote site (if no other nodes at the local site can host the application).
  • The application continues operation using the mirrored copy of the data.
When the initial local site becomes active again, resource group and site management policies determine whether or not the application moves back to the local site:
  • The direction of mirroring may be reversed.
  • The application may be stopped and restarted on another node.
Note: The initial local site may become the new remote site with the direction of mirroring reversed.