Post reply

Disk_Queue_Length is CRITICAL or high

  • April 25, 2013 at 11:58 am

    #276898

    This is regarding the Cluster servers 723/722/721 3 Nodes (Active/Active/Passive). At around Fri Mar 29 11:05:42 EDT 2013, We have got alert from monitoring agent Nagios that “Disk_Queue_Length is CRITICAL” and during that time SQL stopped responding for a couple minutes and this happened about 3 times.

    Since that time I disabled the Nagios alert on that particular event.

    Storage Team/Backup Team/Wintel Team - saying that no issues found from their end.

    Windows Administration team says its everything normal on their side.

    SQL Team - Saying that when this issue happened that time there was no jobs were running on this cluster and no issues.

    Can any one suggest here where exactly issue is and how to resolve this issue.

    Thanks a million in advance.

  • April 25, 2013 at 1:26 pm

    #1610106

    you run anti-virus on those nodes? I always check anti-virus logs when all else appears normal.

  • May 1, 2013 at 2:16 am

    #1611486

    hi Geoff A

    Thanks for your reply,

    I gone through the antivirus log file, but could not find any thing related to the mdf or ldf files.

    will there be any other issues ?, because we disabled the monitoring alert on the cluster nodes regarding this issue,

    I think, I should go for Microsoft support.

    Thanks alot.

  • May 1, 2013 at 2:37 am

    #1611491

    Before you call CSS, I would check the underlying SAN LUNs and be sure there isn't contention there. The SAN admins often look at averages, but if they've carved up physical disks and shared them among LUNs, there can be some contention from other systems.

    No A/V is good, have you also checked if some type of backup or snapshotting was occurring at this time? Anything that might have frozen disk access briefly for SQL Server?

  • May 2, 2013 at 6:40 am

    #1611868

    There are quite a number of things that could be at fault here, especially given the clustering setup you have. I would review every error log you can get your hands on before going to external support. I believe you can get better and MUCH less expensive help than Microsoft CSS too, but I am a tad biased on that statement. :hehe:

    Best,
    Kevin G. Boles
    SQL Server Consultant
    SQL MVP 2007-2012
    TheSQLGuru on googles mail service

  • May 2, 2013 at 9:48 am

    #1611938

    Can you check your error log for IO take longer duration warning message?

Viewing 6 posts - 1 through 5 (of 5 total)

You must be logged in to reply to this topic. Login to reply