July 19, 2007 at 12:37 pm
We have been working on an issue backing up our SQL databases to the network.
To begin with, they work fine locally, and then I can copy them to the network. We have 100's of databases, so this is not practical.
If I run the same jobs to the network. A job that would take 30 to 45 minutes.... will run for hours!!!! Like 6, 8, 13 hours... and sometimes never complete. We end up cancelling the job. Prior to cancelling, if I look to where we are backing up to it appears that the whole file is there....
what could cause this freeze?
don
July 19, 2007 at 1:01 pm
I'm not sure why it's not practical to copy them to the network? Surely you can just script this.
The network latencies sometimes wreck havoc with the SQL backup routines. It expects to be able to stream the pages out very, very quickly.
July 19, 2007 at 1:17 pm
Hey Steve,
thanks for getting back...
We run all the jobs this way and they work fine, and have for years. The issue with these jobs seems to be that the DB size is now 4GB +.
I read that we could disable "caching" on the network interface, because these backups are "cached" prior to being sent over the network. Does this sound familiar?
I realize that we could backup locally.... but we are concerned to with running out of space on the local machine....
Any other options for us?
thanks,
don
July 19, 2007 at 4:36 pm
Not that I know of. I'd add more local space, you need it and don't want to run out. I used to script a job that would just send everything in the backup folder off to a network share every day and let the tape machine gra it from there.
I'd also then delete that stuff before the next backup ran if I had space issues.
Viewing 4 posts - 1 through 4 (of 4 total)
You must be logged in to reply to this topic. Login to reply