Right.  If you want to do hot backups you should use RMAN (Netbackup
type Oracle - requires NB Database Extensions) rather than filesystem
(Netbackup type Standard).

In my case however I was doing a cold backup of an OCFS filesystem.  In
fact the file size (from the OS standpoint) had NOT changed but I've
found other quirks with OCFS.   As an FYI Oracle and Netbackup only
support RMAN backups of OCFS.   We were doing our backup prior to
setting up RMAN and wanted something to fall back on in case our restore
testing in RMAN went horribly wrong.   RMAN has worked like a champ
though.

-----Original Message-----
From: Piszcz, Justin [mailto:jpiszcz AT servervault DOT com] 
Sent: Tuesday, August 16, 2005 9:04 AM
To: Jeff Lightner; Ellwood, MW (Mike); veritas-bu AT eng.auburn DOT edu
Subject: RE: [Veritas-bu] Bpbackup - file "shrunk by nnnn bytes, padding
with zeroes"

You should not be backing up live oracle databases.

-----Original Message-----
From: veritas-bu-admin AT mailman.eng.auburn DOT edu
[mailto:veritas-bu-admin AT mailman.eng.auburn DOT edu] On Behalf Of Jeff
Lightner
Sent: Tuesday, August 16, 2005 8:51 AM
To: Ellwood, MW (Mike); veritas-bu AT eng.auburn DOT edu
Subject: RE: [Veritas-bu] Bpbackup - file "shrunk by nnnn bytes, padding
with zeroes"

It means the file size changed after Netbackup selected it for backup
but before it actually backed it up.

I just ran into this using my Netbackup 5.1 a week or so ago.

-----Original Message-----
From: veritas-bu-admin AT mailman.eng.auburn DOT edu
[mailto:veritas-bu-admin AT mailman.eng.auburn DOT edu] On Behalf Of Ellwood,
MW (Mike)
Sent: Tuesday, August 16, 2005 7:54 AM
To: veritas-bu AT eng.auburn DOT edu
Subject: [Veritas-bu] Bpbackup - file "shrunk by nnnn bytes, padding
with zeroes"

Anyone know what this means?

22:10:13 ERR - File /ora/prod/oradata2/data/bud_tab_02.dbf shrunk by
199754240 bytes, padding with zeros

This is with bpbackup for 3.4GA (old, I know ...).

Oddly enough we were running a parallel backup on another server of the
same data (we do this all the time and it works fine),
 which ran perfectly.

We've been having some tape issues (this is LTO-1 in a Sun-branded L20
robot) on the server where we got the above message.
There is no direct evidence to finger the tape system here, as far as I
can see (nothing in /var/adm/messages, etc).

This is on a Sun V880 platform running Solaris 5.8.

With thanks,


Mike Ellwood


_______________________________________________
Veritas-bu maillist  -  Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu

_______________________________________________
Veritas-bu maillist  -  Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu