Environment
Novell NetWare 6.5 Support Pack 8
Novell Cluster Services 1.8.4
Novell Cluster Services 1.8.4
Situation
The server will be in high utilization with an "NSS WorkToDo" as the busiest thread and/or the message "A scheduled WorkToDo took longer than expected". Migrating the pool to other nodes does not correct the option. Deactivating the pool does allow the CPU utilization to drop back down to normal levels. Running a pool rebuild, however, does not correct the problem.
Resolution
The pool is corrupted. Backup the data, recreate the pool and restore the backup.
Additional Information
The NSS WorkToDo threads are responsible for flushing data to the disk. It appears that the pool is in a state where this flushing process is not able to work properly.