2 probs w/ backup.sh: "Device busy" and "dangling vnode"

[Chad Whitacre]

Dear All,

I am experiencing two problems with the following backup script:

     #!/bin/sh

     /sbin/mount /backup/
     /usr/local/bin/rsync -ax --delete /usr/ /backup/
     /bin/sleep 15
     /sbin/umount /backup/

     echo 'backup of IWS complete'


Here is the cron entry that calls it:

     0       3       *       *       *       time -h /root/backup.sh


Over the past 3 months, this script has yielded the following error 
perhaps once every two weeks or so:

     umount: unmount of /backup failed: Device busy
     backup of IWS complete
            1m28.42s real           2.60s user              6.76s sys


As a result, I upped the sleep time from 5 to 15 seconds 2 days ago. 
This morning, the script seemingly caused the system to crash, given the 
following evidence from /var/log/messages:

     Mar  2 20:11:48 www2 ntpd[442]: kernel time sync enabled 2001
     Mar  2 21:20:08 www2 ntpd[442]: kernel time sync enabled 6001
     Mar  2 23:02:33 www2 ntpd[442]: kernel time sync enabled 2001
     Mar  2 23:36:43 www2 ntpd[442]: kernel time sync enabled 6001
     Mar  2 23:53:47 www2 ntpd[442]: kernel time sync enabled 2001
     Mar  3 01:19:04 www2 ntpd[442]: kernel time sync enabled 6001
     Mar  3 01:36:07 www2 ntpd[442]: kernel time sync enabled 2001
     Mar  3 02:10:19 www2 ntpd[442]: kernel time sync enabled 6001
     Mar  3 02:27:23 www2 ntpd[442]: kernel time sync enabled 2001
     Mar  3 08:18:30 www2 syslogd: kernel boot file is
         /boot/kernel/kernel
     Mar  3 08:18:30 www2 kernel: panic: unmount: dangling vnode
     Mar  3 08:18:30 www2 kernel: Uptime: 55d11h50m47s
     Mar  3 08:18:30 www2 kernel: Dumping 511 MB (2 chunks)
     Mar  3 08:18:30 www2 kernel: chunk 0: 1MB (159 pages) (CTRL-C to
         abort)
     Mar  3 08:18:30 www2 kernel: Dump aborted
     Mar  3 08:18:30 www2 kernel: Automatic reboot in 15 seconds - press
         a key on the console to abort
     Mar  3 08:18:30 www2 kernel: --> Press a key on the console to
         reboot,
     Mar  3 08:18:30 www2 kernel: --> or switch off the system now.
     Mar  3 08:18:30 www2 kernel: Rebooting...
     Mar  3 08:18:30 www2 kernel: Copyright (c) 1992-2005 The FreeBSD
         Project.
     Mar  3 08:18:30 www2 kernel: Copyright (c) 1979, 1980, 1983, 1986,
         1988, 1989, 1991, 1992, 1993, 1994
     Mar  3 08:18:30 www2 kernel: The Regents of the University of
         California. All rights reserved.
     Mar  3 08:18:30 www2 kernel: FreeBSD 6.0-RELEASE #0: Thu Nov  3
         09:36:13 UTC 2005
     Mar  3 08:18:30 www2 kernel:
         root at x64.samsco.home:/usr/obj/usr/src/sys/GENERIC


As you can see, the machine apparently died sometime this morning after 
2:27. The front-line sysadmin (not myself) rebooted the machine this 
morning at 8:18, and reports that at the "chunk 0: 1MB (159 pages) 
(CTRL-C to abort)" message, the system seemed to hang, with no apparent 
hard drive or other activity. He hit CTRL-C, receiving the "Automatic 
reboot in 15 seconds" message, but then the system seemed to hang again, 
at which point he hit CTRL-C a second time, and the machine rebooted.


Any advice on debugging these two problems?

Thank you.




Chad Whitacre
http://www.zetadev.com/