Error in Production backup, please help

Hi All
I'm running Oracle 8i on linux 5.2 .A daily OS backup on the production server is performed. An email is send to every one in the pool. Day before yesterday i got a mail saying that backup is failed. Server is located at a remote place.(NJ- USA. Following is the error message. Can any one look into the error message and throw some light on this. I couldn't figure out the exact cause for this.

ERROR MESSAGE:
----------------
Wed Mar 19 01:00:01 EST 2003: hostdump.pl started

Parameters:
-----------
backuplevel 0
tapedevice /dev/nst0
ejecttape YES
directories / /boot /opt /home /u02
timeout 7200 seconds
-----------

Wed Mar 19 01:00:01 EST 2003: Dump of / started
Wed Mar 19 01:02:21 EST 2003: Dump of / had errors.
Wed Mar 19 01:02:21 EST 2003: The output from this command is appended at the end Wed Mar 19 01:02:21 EST 2003: Quitting Wed Mar 19 01:02:21 EST 2003: hostdump.pl terminating with errors. Tape not ejected

==================================

OUTPUT FROM LAST OS COMMAND EXECUTED

DUMP: Date of this level 0 dump: Wed Mar 19 01:00:01 2003
DUMP: Date of last level 0 dump: the epoch
DUMP: Dumping /dev/cciss/c0d0p10 (/) to /dev/nst0
DUMP: Label: /
DUMP: mapping (Pass I) [regular files]
DUMP: mapping (Pass II) [directories]
DUMP: estimated 1126537 tape blocks.
DUMP: Volume 1 started at: Wed Mar 19 01:00:04 2003
DUMP: dumping (Pass III) [directories]
DUMP: dumping (Pass IV) [regular files]
DUMP: error writing command pipe: No buffer space available
DUMP: The ENTIRE dump is aborted.


Many thanks in advance....

Obviously, this is a home-grown script to backup your filesystem. Without knowing what this script does, how do you expect us to debug it?

I suggest you start reading http://download-west.oracle.com/docs...a96519/toc.htm to understand why taking a straight dump of the disks is not advisable.