Click here to monitor SSC
SQLServerCentral is supported by Red Gate Software Ltd.
 
Log in  ::  Register  ::  Not logged in
 
 
 
        
Home       Members    Calendar    Who's On


Add to briefcase

Cause of Database Mirroring automatic failover Expand / Collapse
Author
Message
Posted Wednesday, June 19, 2013 7:31 AM
SSC Journeyman

SSC JourneymanSSC JourneymanSSC JourneymanSSC JourneymanSSC JourneymanSSC JourneymanSSC JourneymanSSC Journeyman

Group: General Forum Members
Last Login: Yesterday @ 9:32 AM
Points: 94, Visits: 755
Hi,

Today morning our prod database has been failed over from principal to mirror....Please let me know what might be the reason....

Here are the errors..

Principal server error log :

2013-06-19 01:26:27.180 spid27s SQL Server has encountered 1 occurrence(s) of cachestore flush for the 'Object Plans' cachestore (part of plan cache) due to some database maintenance or reconfigure operations.

Witness server log :


6/19/2013 1:30:18 AM, SQL Server Error Log on xxxxxxx is Warning.

SQL Server instance xxxxxx recorded 1 error log message requiring attention.

The message are shown below in context (if applicable):
Warning 6/19/2013 1:26:20 AM spid10s Error: 1479, Severity: 16, State: 1.
OK 6/19/2013 1:26:20 AM spid10s The mirroring connection to "TCP://xxxx.ds.xxxx.com:9030" has timed out for database "xxxxxx" after 10 seconds without a response. Check the service and network connections.

environment :

Sql server 2005 SP3 standard edition
windows server 2003
Database mirroring configured with HA

Please let me know if you need more details....
Post #1465160
Posted Wednesday, June 19, 2013 8:28 AM
Right there with Babe

Right there with BabeRight there with BabeRight there with BabeRight there with BabeRight there with BabeRight there with BabeRight there with BabeRight there with Babe

Group: General Forum Members
Last Login: Friday, August 8, 2014 4:15 PM
Points: 727, Visits: 1,417
Just like the error states, the witness couldn't reach the principal for 10 seconds.

This could happen for any number of reasons. An extremely high work load that causes the server to be come unavailable, a network disruption lasting longer than 10 seconds, drop network packages, vm issues, DNS issue, etc.

In my past life, we had the witness on a vm, which was over provisioned. So when the server physical server running the VM went into overdrive, it started causing issues. As a temporary fix, we increased the duration of the mirroring timeout from 10 to 30. Once we had a new machine, we moved the VM, which resolved a number of the lost packets and sudden connection issues that we saw.

Fraggle.
Post #1465209
Posted Wednesday, June 19, 2013 9:39 AM
SSC Journeyman

SSC JourneymanSSC JourneymanSSC JourneymanSSC JourneymanSSC JourneymanSSC JourneymanSSC JourneymanSSC Journeyman

Group: General Forum Members
Last Login: Yesterday @ 9:32 AM
Points: 94, Visits: 755
thanks for the response....
but i dont see much activity in the instance...and it occured around 1:20 AM....
Post #1465255
Posted Wednesday, June 19, 2013 10:04 PM


Old Hand

Old HandOld HandOld HandOld HandOld HandOld HandOld HandOld Hand

Group: General Forum Members
Last Login: Friday, March 21, 2014 9:46 AM
Points: 387, Visits: 1,078
Robin35 (6/19/2013)
Principal server error log :

2013-06-19 01:26:27.180 spid27s SQL Server has encountered 1 occurrence(s) of cachestore flush for the 'Object Plans' cachestore (part of plan cache) due to some database maintenance or reconfigure operations.


proccache has been flushed/remove. It could because of server / database level configuration change or memory pressure.

Some times, if principal is running with low resources like cpu,memory (in your case of memory pressere) there is a possiblity of principal's low response to witness that may do failover.
Post #1465449
« Prev Topic | Next Topic »

Add to briefcase

Permissions Expand / Collapse