Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations Chris Miller on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

NTBackup sometimes failed

Status
Not open for further replies.

ludolf2

Technical User
May 11, 2009
7
HU
hi all

We have a cluster environment with two server2003 ent sp2 node (active/passive).
The scheduled task service is clustered.
The daily ntbackup task regularly finished with a bunch of error message like this:
WARNING: Portions of "[filename]" cannot be read. The backed up data is corrupt or incomplete.
This is the only error type.
The backup fails on the storage disks and also on the node's boot disk. Storage disk are in raid5, the system disk is mirrored.
Chkdsk didn't found any errors.
Tried both node.
The "unreadable" files are random, the user which we running with the ntbackup can read the files.
The "vssadmin list writers" comman show "Last error: No error" for all writers.
The KB940349 hotfix is installed on both nodes.
The systemstate backup is scheduled in the next form: a batch file executes ntbackup on both node with psexec and unc path (\\server). This backup is always finished successfully on the inactive node, and always fails on the active...
Could someone help me?
 
on the active backup, are you using the virtual servername to perform those backups?

_______________________________________
Great knowledge can be obtained by mastering the Google algorithm.
 
on the active node we are using the local disk drive letters.
for example backup S: to X:

 
If it is the Quorum and S:(shared data drive) you are after, you will need to set it up as a network share backup. Run the NTBackup from another server that has S: and Q: mapped and back them up that way. Otherwise these may help. The trick is, you must use the Virtual Cluster Name (Mail Server name everyone uses) because you never know which node will be active.

How to Back Up the Exchange 2003 Databases That Are Stored on Shared Disk Resources
Backing Up an Exchange Server 2003 Cluster's Shared Disk Resources


_______________________________________
Great knowledge can be obtained by mastering the Google algorithm.
 
thanks for the answer, but if we run ntbackup on mapped drives, opened files will not backed up (vss only works on local hard drives)
 
its most likely due to the cluster service having a lock on the volumes while being the active node. to meet your goals, you likely need to go with backing up the inactive cluster node, then performing a failover from the active to inactive node, then performing the backup against that node.

- Brandon Wilson
MCSE:Security00/03; MCSA:Security03
MCSA:Messaging00; MCP; A+
IT Pangaea (
 
I've tried this before. Ntbackup fails on both node.
Additional info: I've just noticed that in the "vssadmin list writers" list the status for almost all writer is "waiting for completion". What does it mean?
 
Since you are running a cluster, I would assume that these servers are very important...:) I would look at other backup software to perform a proper backup. ntbackup is best for one time backups as needed, or non-critical servers. A third party backup software can backup from another machine using VSS, and therefore collect all needed data.

Just my 2 cents. :)

_______________________________________
Great knowledge can be obtained by mastering the Google algorithm.
 
Yes, those are important servers (DC,DHCP,DNS,SMS,WSUS...)
I think in this case the problem is in the vss, not the backup program.
As in the subject, the backup is sometimes complete without errors, and other day is incomplete without any configuration change.
Could you tell me, what does "waiting for completion" means?
 
I've checked the file versions: our files dated 2007.02, 2007.08, and 2007.09.
Some hours ago i noticed that a rsmsink process stucked on the server, running in the backup user's context. Killed it and after this I made a backup interactively, and it finished without errors. I will report the next scheduled backup's results.
 
that is the removeable storage service, and it should be running when a backup is kicked off....

ikt looks like that error may be able to be resolved via a hotifx potentially...check:



- Brandon Wilson
MCSE:Security00/03; MCSA:Security03
MCSA:Messaging00; MCP; A+
IT Pangaea (
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top