September 6, 2005 at 10:21 am
Hoping you gurus could help me out.
I have a cluster with two identical active/passive nodes.
However over the past week i've noticed the backups taking 15 times longer than normal, thus hanging the trans log backups as well.
After checking the server, memory, space, cpu i decided to transfer the group to the second node while rebooting the first.
As the first node was rebooting, i decided to backup a database to see if there was any performance change, and there was!
the transfer rate on node1 was .493 MB/sec
the transfer rate on node2 was 18 MB/sec
even though these machines are identical and writing backups to the same local drive what could be causing this?
Any ideas would be greatly appreciated.
Thanks,
Warren
September 6, 2005 at 10:43 am
What are you backing up to and how is it configured to the cluster?
When I had an active/passive cluster, I backed up to a tape library. The tape library was connected, via the fabric, to node 2. Backing up via node 1 was slow while via node 2 it was very quick.
Check which node 'owns' the tape library/device.
-SQLBill
September 6, 2005 at 10:50 am
I'm backing up the databases locally to the shared shelf only.
I'm using a seperate server to copy the backups to a standby along with backing up to tape.
so the clusters are only backing up to a local drive.
September 7, 2005 at 8:45 am
Is the backup disk in the same cluster group as the sql server?
do you access the disk via drive letter or via network share?
sql server native backup or third party tool?
are master, tempdb and msdb on clustered disks or on local disks(->same config for server service)?
are both nodes identical (memory, cpu's, hyperthreading, pagefile?
All that (and more) might have an impact on backup times...
regards
karl
Best regards
karl
Viewing 4 posts - 1 through 3 (of 3 total)
You must be logged in to reply to this topic. Login to reply