May 28, 2014 at 10:57 am
I am seeing the Sch-S + Sch-M deadlocks on our system. They occur because the maintenance online index rebuilds runs into a user/batch process. It seems like this should be pretty rare being that the Sch-M is only held briefly but I guess the high level of activity combined with lots of indexes causes issues.
Do you think I would have better luck with index reorganize? I profiled both and index reorg does a lot less locking. I could just save rebuilds for the weekend and such.
Thank You,
Scott
May 28, 2014 at 12:07 pm
Reorganizing indexes will generate a lot more transaction log activity, as each row moved will need an entry. For highly fragmented tables, it will also take more time.
May 28, 2014 at 1:19 pm
Ugh. Thanks, will keep this in mind.
May 28, 2014 at 7:33 pm
How big is the table?
--Jeff Moden
Change is inevitable... Change for the better is not.
May 28, 2014 at 8:02 pm
Here is an interesting article on the effects of a rebuild vs. reorg. The effects of the article seem pretty consistent whether clustering on a GUID or Int and with or without compression enabled.
Jason...AKA CirqueDeSQLeil
_______________________________________________
I have given a name to my pain...MCM SQL Server, MVP
SQL RNNR
Posting Performance Based Questions - Gail Shaw[/url]
Learn Extended Events
Viewing 5 posts - 1 through 4 (of 4 total)
You must be logged in to reply to this topic. Login to reply