1 Node in Filestore cluster crashes
|Article:TECH189403|||||Created: 2012-05-23|||||Updated: 2012-05-23|||||Article URL http://www.symantec.com/docs/TECH189403|
A node in a FileStore cluster faults for no reason. The node will be in a shutdown state and require rebooting.
The messages file on the node will show lines similar to the following:
2012 Apr 16 14:25:25 cluster_02 [err] kernel: [10462.770410] update_cpu_power: cpu_power = 10577420574567; scale_rt = 18428740252620644199
N8300 FileStore appliance. Versions below 5.7P3.
The issue is caused by a bug within the kernel used by the N8300 appliance. After 208 days of uptime an internal timer is incorrectly calculated and a buffer overrun can occur, causing the server to crash.
The issue will be fixed in 5.7P3 patch.
As the issue only occurs after 208 days of uptime, as a workaround the server can be rebooted prior to this to prevent the crash.
Article URL http://www.symantec.com/docs/TECH189403