Sorry if it has been discussed before, but I can't find a definite answer for the simple task of bringing down a cluster safely. Suppose I have a cluster with 4 nodes: 2 is in a HA configuration using DRBD, working fencing devices and config, and 2 other nodes for various purposes. All I want is to safely power down all the nodes in the cluster, without triggering fencing, automatic relocation of VMs by failover migration, etc. By hand, or by automation/scripting started by eg. an UPS monitoring software.
How to do it?
How to do it?