When one gluster host is shut down, VM’s running on other gluster hosts gets killed.
Issue
-
When one gluster host 'A' is restarted, then VMs running on other hosts (for example host 'B' ) were killed because there was VM lease set on gluster storage and gluster storage was unavailable on other hosts (host 'B') when first host 'A' was restarted.
-
Here, for example, when we restarted host 'rh5' from iLO console and after that VMs running on host 'rh2' were killed due to sanlock error. '
Host 'rh5' restarted on May 18 11:38:
=========
May 18 11:38:44 test-rh5 systemd-logind: Power key pressed.
May 18 11:38:44 test-rh5 systemd-logind: Powering Off...
=========
Environment
- Red Hat Hyperconverged Infrastructure 1.5
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.