Server :: Can't Kill A Backup Job - Space Becoming Limited

### update ####

Sorry to waste the groups time on this one, after killing that and biting fingers, I re-examined the ps list and saw the tar and bzip running fine (so even though I killed the backup .sh, it was still going along, so I simply removed those and all was well again.

Feel free to reply though with #3, if I have 30G of mail, I know gzip is faster but bzip is more compression from reading, should;
a) tarring that mail actually drop to 3G total!
b) est (I know it's tough) but backing up 30G, is 10 hours longer than expected? (I will run some tests on a smaller folder)

############### end update

this is on an old RH9 box and backing up mail. I started the job last night in a shell script, it's around 30G of mail, and it was a tar using bzip as there was only 20 or so gig 'free space'. The old backup script was in perl and just a tar mail folder and the total was 3G so I figured I was safe.

Well it started last night and this am it was still running. I did a kill -9 on the pid and ps now shows it as;
root 9143 0.0 0.0 0 0 ? Z Feb23 0:00 [backup.sh <defunct>]


Disk space was down to 1.3G free (so somewhere all that space is being held temp somewhere), but I removed the old backup (3G which gave me a little breathing room of 4.5G), but the backup is still running somehow as the 4.5 is now 4.3. I tried the pkill, kill -9 pid, read someone said re-start the job which I did, then killed but nothing.

I really can't reboot this production box, so in RH9, I need to both;
1.kill that defunct backup
2.remove the temp storage it has made
3.figure out why it's taken 10+ hours and not done.

1 and 2 being 'critical'