On Nov 16, 2004, at 3:31 AM, Bos, Karel wrote:

I am looking for the reason for getting this message in the error.log:

11/16/2004 09:23:17 CreateSnapshotSet(): pAsync->QueryStatus() returns
hr=E_OUTOFMEMORY
11/16/2004 09:23:29 ANS1999E Incremental processing of
'\\clusternode3\c$' stopped.

11/16/2004 09:23:29 ANS1950E Backup via Microsoft Volume Shadow Copy
failed.  See error log for more detail.

I have 4 cluster Win2003 cluster nodes configured with the same
dsm.opt and client version. Only one of them gives this message. All
of them have the same amount of CPU power and memory. The one failing
is the least used.

One other thing, after a reboot the first back up will run fine. All
thereafter fail.

Well, you have run out of virtual memory in that system, either because
of
outright exhaustion or memory fragmentation such that a standard block
of
memory cannot be allocated from the heap.
Windows Service Packs sometimes fix memory management problems.
Boosting the heap size in the Registry may be an approach, but I would
not go there until other things have been examined.
Use the Task Manager or like facility to see what's eating memory.
If you are running backups via the Scheduler, instead consider using the
Client Acceptor Service such that memory is not tied up.

    my $0.02   Richard Sims