Monday, March 12, 2012

Reindexing and Transaction Logs

Hi All,

Just after some feedback on a scenario where we have full logging setup on one of the databases, and the transaction logs are backed up every 60min. At 0000-0100 the log jumps from being a few thousand k up to over 1.7gb.
I did some profiling for this time, and it appears that this jump is related to the reindexing of the indexes on the database.

Is this normal for the log file to jump in size so much? Or is this an indication of some other issue (potentially with the indexes)?

Is there any way that the reindexing can be excluded from the log files or is this a necessity?

Thanks in advance for your help.

Cheers
TroyIt is quite normal for re-indexing to fill up a transaction log. Especially if you use dbcc indexdefrag.|||So I presume there is no 2 ways around this really then? Massive jump in size, which makes for some interesting issues in terms of planning for space *sigh*...|||Your log may stay smaller with database recovery model set to SIMPLE|||haha thanks for that - umm but then I loose my logging...sorta defeats the point of backing up the logs :P
But yeah I hear what you are all saying - suppose I was just after a solution which would allow me to turn off the logging for the index rebuilds and leave it on for all the remainder of the normal i/u/d's in the database...

Thanks all, have a good Christmas/Break/etc...

No comments:

Post a Comment