When Cluster Service detects a corrupt Quorum log, it will automatically attempt to repair it. This is indicated by an entry in Event Viewer. If Cluster Service cannot repair the log, the service will not start and the log must be manually reset or repaired. Also, if Cluster Service fails to detect a corrupt quorum log, an ERROR_CLUSTERLOG_CORRUPT message will appear in the Cluster log file.
You can attempt to repair a corrupt Quorum log by following the steps described below:
- From the command prompt, switch to the Cluster directory and type clussvc -debug -resetquorumlog.
- Stop and restart Cluster Service using the net stop/net start commands. If the log file cannot be reset, start Cluster Service using the -noquorumlogging parameter.
- Start the Cluster Service on one node with the -noquorumlogging parameter entered in the Startup Parameter field from the service’s properties window.
- Run chkdsk on the Quorum disk. Of chkdsk finds errors, elect to repair them. If no errors are found, delete the quolog.log file and any temporary files located in the MSCS directory on the Quorum disk.
- Stop and restart the Cluster Service on the one node.
Once you deleted the quolog.log file, any configuration changes that were in the log and not communicated to all nodes will be lost. However, because the log is corrupt, that information is irretrievable, and this is the only course of action to take.