Massive database growth

  • I have a database we'll just call someDB.

    someDB, over the past five years, has gone from about 1.9 gb to about 2.5 gb. That is until about a week ago, when someDB ballooned to 10 gb, a four-fold increase.

    someDB is in full recovery, but the log file size is the same (roughly) as it was before the sudden increase.

    Free space is sitting at < 300 mb. One might think that there was a sudden influx of data, but one would be wrong. I'll focus on one particularly egregious table as an example. Table: case_action Before the Bloat:
    DBCC results for 'case_action'.
    There are 1815153 rows in 61317 pages for object "case_action".

    reserved1,106,872 KB
    data 620,992 KB
    index_size 483,656 KB
    unused 2,224 KB

    After the Bloat:
    DBCC results for 'case_action'.
    There are 1815885 rows in 222160 pages for object "case_action".

    reserved3,800,216 KB
    data1,907,840 KB
    index_size1,889,552 KB
    unused 2,824 KB

    Change from Bloat:
    ROWS 732
    reserved2,693,344 KB
    data1,286,848 KB
    index_size1,405,896 KB
    unused 600 KB

    I can't figure out what happened to increase the database size. I have restored the database to a test server in both it's pre- and post- bloat forms and tried running the couple of things I did to it in the time frame the growth occurred, to no avail. These include rebuilding a couple indexes manually, though that was not on the table outlined above.

    My transaction log backups don't reflect the size change either. The differential backup the night before the growth is the normal size, the transaction log backups throughout the day are the normally expected size, then the next differential is nearly 4 times bigger then the previous one.

    Any suggestions as to what to look at next would be greatly appreciated.

    Here's hoping I'm missing something simple.
    Steve.

  • Has the fill factor changed on your index and a rebuild index has been performed??

  • What is the fill factor on the table?

  • Funny you should mention that. I just scripted out the creates for the table above, dropped them into text files, and used beyond compare to check them.

    Guess what popped up as the only difference...

    , FILLFACTOR = 25

    Imagine that.

    Sometimes I just gotta talk something through before the big stick comes round for the head smack.

    Thanks,

    Steve.

  • Steve Hoyer (8/12/2009)


    Imagine that.

    Got any idea who the last person to rebuild indexes before the growth was?

    Gail Shaw
    Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
    SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability

    We walk in the dark places no others will enter
    We stand on the bridge and no one may pass
  • Nope. No Idea. This is not the moron you're looking for.

  • Steve Hoyer (8/12/2009)


    This is not the moron you're looking for.

    I was not implying it was you. Sorry if you got that impression.

    Gail Shaw
    Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
    SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability

    We walk in the dark places no others will enter
    We stand on the bridge and no one may pass
  • Naw. You see that was merely my clever why of deflecting well deserved blame. I'd never actually admit to anything like that.

    Wait.

    Crap.

  • Hah! Looks like a Rookie used the Force on the Mighty Gila! 😉

  • There are four ways of memory extension ie /3GB, /USERVA, /PAE and AWE. Are all the windows technologies or any one of them is sql server technology?

    Post only if you are sure.

    -LK

  • Please post new questions in a new therad. This has absolutely nothing to do with large databases

    luckysql.kinda (8/13/2009)


    There are four ways of memory extension ie /3GB, /USERVA, /PAE and AWE. Are all the windows technologies or any one of them is sql server technology?

    All windows.

    Post only if you are sure.

    I wouldn't post if I wasn't. Do you want supporting references?

    Gail Shaw
    Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
    SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability

    We walk in the dark places no others will enter
    We stand on the bridge and no one may pass

Viewing 11 posts - 1 through 10 (of 10 total)

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