How often (in seconds) a background thread proactively checks that the pooled vSphere session is still alive by issuing a lightweight currentTime() call to the vCenter.
If the call fails the pool reconnects immediately, so the next operation can proceed without waiting for a timeout.
Each health check is itself one vCenter API invocation, so it will show up in the vCenter Events log / API invocation counters even though it is not counted against Pool session max uses below (that setting only counts sessions handed out to Jenkins operations, e.g. clone/power-on/snapshot/delete).
Set to 0 to disable health checks entirely — the pool will still reconnect reactively if an operation fails due to an expired session.