SQL backups failing
All of a sudden some of our SQL backups are failing with this error: Operating system error 64(The specified network name is no longer available.). These backups are being written to a different server, then backed up to disk. To rule out DNS issues, I changed the failing ones to use IP. Still failing. Turned off virus scanner, so it wouldn't scan while the file was being written. Still failing. The only thing that has changed was the applying of the June Security Patches.
Software/Hardware used:
Software/Hardware used:
ASKED:
July 6, 2005 8:49 AM
UPDATED:
July 18, 2005 10:55 AM
Answer Wiki:
What ID is being used to perform the backups?
To see all answers submitted to the Answer Wiki: View Answer History.
The user has SA privileges.
We had 5 jobs fail last night, 2 of which have failed regularly for the past 20 days. The other 3 have run fine. I am able to successfully run the jobs manually.
The question about the ID has to do with the Windows network ID that would be used for the network share. It seems like you have permissions to run the backups on the SQL server machine but probably not permissions to access the network share. When you login as ???? to run the backups you are using a different account to access the network share and thats why it works.
Be sure to post the fix when you figure it out.
– Good Luck
This user is also a member of Domain Admins and is in the local Administrator group on the server where this file is being written.
Is there a space issue on the destination drive? If the succeeding backups were of small databases and the large ones were failing, hard drive space could be the culprit.
Not a space issue. There’s 69.5 GB free space on the server.
Are you using a share name like bakserversharename? If so, could it have been removed?
The UNC path does exist and the backups actually start just fine. The error messages occur after the backups have started.
The cause of this problem may be the size of your backup. See the microsoft article at ‘http://support.microsoft.com/default.aspx?scid=kb;EN-US;q285997′ about how creating snapshots fails due to network directory caching.
The problem you’ve seems a OS problem. So, I think it’s worth to try a absolute file path instead of UNC path.
We are not seeing the errors in the event log that the KB article talks about. The only errors we see are SQLSERVERAGENT and MSSQLSERVER errors on the SQL server.
Last night we changed the path of some of the backups that fail most frequently from a server to our NAS, and they worked. One that we did not change failed again last night, so it must be something to do with the server we are writing to, even though most of the backups work just fine.
I have a similar problem,
since middle of last week, all my maintenance plans with backup (transaction and full) are hanging. The backup file is present, but the job is still active. I have to stop manually the job or kill the process sqlmaint.exe.
When I run a manual backup from the Query Analyser it works.
Did you try a backup from the Query Analyser ?
We are still woking on this issue…
We upgraded the server where the backups are being written to 2003 and have not a single backup fail since.
Thanks for all the help!!