Planning for fail-over scenarios and upgrade

Information on performance impact and management during failover scenario of IBM Storage Scale protocol node.

During the planning of a system that has a high number of active concurrent SMB connections, the potential performance impact during fail-over scenarios needs to be considered.

If an IBM Storage Scale protocol node fails, the IP addresses that are hosted by that protocol node are relocated to other IBM Storage Scale protocol nodes, and SMB client reconnections are redistributed among the remaining IBM Storage Scale nodes.

When a high number of active concurrent SMB connections are planned, the overall system configuration needs to factor some buffer space in terms of maximum active concurrent SMB connections. This accounts for the potential performance implications during these fail-over scenarios.

During the IBM Storage Scale software upgrade process, IP addresses are frequently relocated. Depending on the configuration, multiple protocol nodes are suspended concurrently to minimize the upgrade time. This leaves fewer protocol nodes to serve various protocol clients including SMB. Therefore, the maximum number of active SMB connections cannot be sustained during the IBM Storage Scale software upgrade process. You need to plan for upgrade during the off-peak hours or schedule a maintenance window to minimize the impact on clients that access IBM Storage Scale.

For more information on upgrade planning, see SMB fail-over scenarios and upgrade.