Description of problem:
The customer disconnected an RHS node to perform maintenance, resulting in the 42 second timeout period in RHS for failover to occur. During this hang on the storage, Sanlock reported errors and vdsm crashed on the SPM. Other hosts weren't affected.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. Create RHEV storage domain using RHS storage with 2 nodes.
2. Shut down one of the RHS nodes, triggering the 42 s timeout for failover
SPM reports sanlock errors, and vdsm crashes and respawns, finally resulting in a fencing event
Paused VMs, latency warnings in Events tab, followed by recovery