Token is lost and cman killed on multiple cluster nodes when a time jump occurs in RHEL 5
Issue
- The token was lost in the cluster, both nodes in the two-node cluster attempted to rejoin, and cman on both nodes was killed without either node being fenced.
- System messages show a token loss immediately after ntpd does a time reset.
- Why did the
ntpddaemon cause a shutdown on a cluster built on Red Hat Enterprise Linux 5.6? - Why does my node reboot when ntp causes a large time jump on the system?
Environment
- Red Hat Enterprise Linux Server 5 (with the High Availability Add on)
openaisprior to 0.80.6-28.el5_6.4 or prior to release 0.80.6-30.el5_7.4
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.
