July 25, 2013 at 7:11 am
Sometimes our log shipping restore job takes too long to complete. Though the file sizes(around 1 GB) are small it takes 4-5 hours to finish. The wait type we observe is IO_Completion. This happens once in a while. When i check the percent_complete it shows 100 percent but the job still runs. The lag is above 12 hours. I am trying to restore the differential backup on to the secondary server database to keep the databases in sync without breaking the log shipping. But the restore job continuously picks all the files one by one that are to be restored. If I stop the job the database will go into suspect mode. I had this issue before. I moved all the files which need to be restored to other location.
How to stop the job without disturbing the state of the database so that i can continue with differential backup restore and then go back to log backup restore?
July 25, 2013 at 4:48 pm
What about stopping the copy job or moving the log files waiting to be restored to a separate temp folder on the same drive ? Not ideal but at least the job won't be able to restore any more files - its going into suspect mode because you're stopping a log restore before its complete.
Are you using ms log shipping or your own method ? We use powershell for our log shipping and it works really well - Worth checking this link;
http://www.sqlservercentral.com/articles/powershell/99832/
Cheers
Vultar
July 26, 2013 at 7:49 am
Regarding the log running restore job, you are probably dealing with extensive VLF fragmentation.
Run DBCC LOGINFO against the database in question to check how many VLF's you have
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply