On Jun 14, 12:09 pm, brent.mccrac...@gmail.com wrote:
> I've run out of options so I'm putting it out to the community for a
> solution. The error message is;
>
> BackupMedium::ReportIoError: write failure on backup device
> 'VDI_0AF26571-DF14-4572-88BE-4E2097AA25C5_0'. Operating system error
> 64(The specified network name is no longer available.).
>
> We're backing up the database to a network share. If you dont like
> that then sorry, but that's the way it is. This wasn't a problem in
> the previous platform, so we can rule out the idea of backing up to
> local disk and copying the file across. We're running SQLlitespeed x64
> and the database is 1.1TB and the backup file is just over 300GB. A
> successful run takes about 3.5 hours.
>
> I dont think it is a network error... certainly not connectivity. I've
> thoroughly tested this and everything is intact. This is what I've
> found/checked;
>
> NICS are not dropping packets; Happens on both nodes of the cluster;
> ran a ping test for duration of the backup - all ok. Ran netmon and
> didn't notice anything. So, I think that the network error is generic/
> erroneous.
>
> I did make a change that fixed the Mrxsmb error 50 and that was to
> change the WINS settings from default to "Use NetBIOS over TCP/IP"
>
> This has started occuring since we moved from x86 to x64. New hardware
> and SAN. The environment is win2k3 r2 enterprise and sql2005
> enterprise all native 64 bit.
>
> This is the clue... the problem occurs at the same time... 42mins and
> 30 seconds into the backup. There isn't anything else running that is
> causing this to fail at this time - remember it is intermittent so it
> does run successfully.
>
> I don't think the VSS is causing the problem either.
>
> any ideas?