October 4, 2006 at 8:52 am
Hi
Im doing some testing on mirroring at the moment, and have found out about the inherent problems with transaction log size that thie replication scenario entails. However I have found that frequent trans log backups are a way around the problem - This sets the vast majority of VLFs in the trans log to state 0 allowing them to be overwritten by new transactions VLF's. Accordingly I dont need to autogrow half as often.
The question is which is the most harmful of system performance? frequent trans log backups or autogrows? if i DO backup frequently i DONT need to autogrow, and if i DONT back up frequently i DO need to autogrow...
which which people choose?
Cheers
Alastair Jones
Methodology Group
October 5, 2006 at 10:30 am
autogrow can cause unexpected fragmentation, both internal and external. if you have a maintenance plan that regularly rebuilds indices, and an external program such as Diskeeper that is scheduled to run regularly, then perhaps this fragmentation isn't an issue. in the old days, "real dbas" never used autogrow because they preferred to manage things on their own. I personally have better things to do... and use it all the time.
Thank-you,
David Russell
Any Cloud, Any Database, Oracle since 1982
Viewing 2 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply