Clustered Red Hat Enterprise Linux 5 servers had to be rebooted after dlm_send process hit 100% usage
Issue
In a clustered Red Hat Enterprise Linux 5 environment, disk I/O has been steadily increasing over 8 days or so, culminating in extremely high load and disk I/O contention. After attempting to remove node 1 from the cluster, the dlm_send process on the node started consuming 100% CPU, and required a reboot of the machine to rectify.
Environment
- Red Hat Enterprise Linux 5
- GFS2
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.