Solaris With Grid Infrastructure Home On ZFS: Cluster Nodes Hang When Executing crsctl Commands And System Becomes Un-responsive

(Doc ID 1622825.1)

Last updated on JUNE 05, 2014

Applies to:

Oracle Database - Enterprise Edition - Version 11.2.0.3 and later
Oracle Solaris on SPARC (64-bit)

Symptoms

Clusterware GRID_HOME is installed on Solaris ZFS local file system. Users complain about slow logins to the node when the clusterware is running. Copying a file (cp) to a local ZFS file system takes more than 6 minutes while copying the same file to NFS mounted file system takes only 2.55 seconds. Sometimes, when running 'crsctl check crs' or 'crsctl stat res -t' on the Cluster nodes, the command appears to hang. Oswatcher data shows hardly any load on the system. Frequent unexplained node reboots are also being observed.

The Cluster node alert logs shows that agents timed out prior to the hang:

 

Cause

Sign In with your My Oracle Support account

Don't have a My Oracle Support account? Click to get started

My Oracle Support provides customers with access to over a
Million Knowledge Articles and hundreds of Community platforms