I am having an issue with Sun EBS Networker 7.3, (64 bit) (Sun Hardware, Solaris 9) where it expects an Oracle DB (11.5.10 with 10g Database (10.2.0.2)) size of 4294967295 bytes. The DB’s get backed up anyway, but it’s only happening to certain DB’s. The warning messages occur at random intervals so it’s difficult to find a logical explanation. Below is a sample from the log. Any suggestion would be welcomed. Thanks.

* <WARNING> : Warning - `/path/to/db’s/applsysd01.dbf' size grew during save
* <WARNING> : Expected 4294967295 bytes for `/path/to/db’s /applsysd01.dbf', got 7444897792 bytes

* <WARNING> : Warning - `/path/to/db’s /applsysd02.dbf' size grew during save
* <WARNING> : Expected 4294967295 bytes for `/path/to/db’s /applsysd02.dbf', got 7654612992 bytes

* <WARNING> : Warning - ``/path/to/db’s /temp01.dbf' size grew during save
* <WARNING> : Expected 4294967295 bytes for ``/path/to/db’s /temp01.dbf', got 8193581056 bytes

Seems the database is so large (relative to your I/O processing power) that the time it takes to back it up is so long that the database is updated in a way that affects the byte check, doesn't it?

Isn't this normal behavior with large DBs constrained by disk I/O?

The DB's aren't up while the backup is running. We have a shell script stopping all activity to make sure the backup is clean.

Unfortunately, this isn't normal behavior in that case we might have more ideas. It occurs even on the same backup job randomly. This same job ran two days prior and succeeded without incident.

We’ve contact Sun regarding a possible bug in the software, but have heard nothing back.

Check sunsolve document id: 209336

Solution 209336 : Sun StorEdge[TM] Enterprise Backup Server (EBS 7.3): Warning - size grew during save

Example:
Dec 2 09:32:43 epbup root: [ID 702911 daemon.notice] * epbup:/ save: Warning - `/usr/local/samba/var/log/log.nmbd' size grew during save
Dec 2 09:32:43 epbup root: [ID 702911 daemon.notice] * epbup:/ save: Expected 1745449 bytes for `/usr/local/samba/var/log/log.nmbd', got 1745828 bytes

Cause:
The backup server detected that the number of bytes associated with the file
on the backup media is not the same size as in the original file. This
inconsistency might occur if the specified file:
? Is a Hierarchical Storage Management (HSM) stub or sparse file.
? Changed during the backup process, and its file size increased.
? Is the media database. Multiple groups may have completed their backup
sessions simultaneously. These sessions were subsequently attempting to
update the media database simultaneously.

You can check the file status using:

#ls -lu
#fuser

which will check when the files were last touched and if they are
still in use

Then as appropriate perform an ad hoc backup of the specified
file to ensure that a consistent copy of the file is backed up.

Thanks for the suggestion. Our error is more specific to the file size being larger thank 2GB, and the file changing when it actually hasn't. Since it's still backing up the DB's it might be something we can live with.