pulse sometimes fails to add all configured virtual IP addresses when service is started or failed over from other piranha host
Issue
- When stopping the pulse service on the active piranha LVS director host, not all of the virtual IP addresses ("VIP's") come up successfully on the standby piranha LVS director host
- Problem occurs whether the active host is the primary or the backup host
- Problem occurs "in both directions" -- when stopping the primary, the backup does not always activate all virtual IP's. When the backup is active and stopped, the primary does not always activate all VIP's either
- Manually restarting the pulse service on the new active piranha host causes all VIP's to come up
- Problem can also occur with a single piranha host (no standby/backup configured)
Environment
-
Red Hat Enterprise Linux 5 Update 6
-
Red Hat Cluster Suite
- piranha
- multiple virtual IP addresses configured as services
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.