In an active-passive cluster, which node should be patched first if you want to minimize service disruption?

In an active-passive cluster configuration, one node is actively handling traffic and workload (the active node) while the other node remains on standby (the passive node) and does not process requests unless it is needed to take over. To minimize service disruption during patching, the passive node should be patched first.

By patching the passive node, it remains ready to take over in case any issues arise during the patching of the active node. Once the passive node is successfully patched and verified to be functioning correctly, the active node can then be patched. This approach ensures that if something goes wrong during the patching process of the active node, the passive node is fully operational to take over, maintaining continuity of service.

Patching the active node first could lead to service disruption if the patch causes issues, and if both nodes are patched simultaneously, there is a risk of both nodes being unable to handle requests at the same time. Hence, patching the passive node first is the strategy that allows for the highest availability and least chance of service disruption during maintenance activities.