RE: Q's about IBM TSM (was Re: HEADSUP: ibcs2 and svr4 comp at headed for history)
From: Paul Mather (paul_at_gromit.dlib.vt.edu)
Date: 06/30/04
- Previous message: Poul-Henning Kamp: "Re: TEST/REVIEW: minor improvement to linux-o-lator"
- In reply to: Thyer, Matthew: "RE: Q's about IBM TSM (was Re: HEADSUP: ibcs2 and svr4 comp at headed for history)"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
To: "Thyer, Matthew" <Matthew.Thyer@dsto.defence.gov.au> Date: Wed, 30 Jun 2004 09:30:18 -0400
On Wed, 2004-06-30 at 00:45, Thyer, Matthew wrote:
> I would love to have this automatic generation/maintenance of /compat/linux/etc/mtab.
>
> This would help me greatly for backup of systems using Linux backup software.
>
> Currently the software (HP OpenView DataProtector) will fail to backup the system if the file does not contain lines regarding the file systems I want to backup.
>
> Ideally this would need some kernel support to handle updating of the file whenever file systems are manually mounted and unmounted. Remember that Linux's etc/mtab file is to record what is currently mounted. A simple translation of FreeBSD's /etc/fstab at module load time or system boot time is not really adequate.
In my case (using Tivoli TSM to back up snapshots via the TSM scheduler)
I use the "preschedulecmd" script hooks to update the mtab file to
reflect the current mounts (and make the snapshots). If your backup
client has pre- and post-backup scripting support, you could take
advantage of this approach to modify mtab appropriately.
Another alternative would be to make a wrapper script around your backup
client that updated mtab and then called the real backup client.
Unless your mounts are very volatile, even a cron job that updated mtab
before the anticipated backup window (or a manual script execution that
did the same prior to a manual backup) would suffice in the absence of
any automagic kernel support.
Also, it's unclear what the correct semantics for an automatic
kernel-based translation would be. In the case of Tivoli TSM, you have
to lie and say your ufs filesystems are ext2, because TSM doesn't
understand or recognise ufs but it does ext2. It's not clear whether
such a lie is valid for all programs run under Linux emulation.
However, performing a "straight" translation, preserving the true
filesystem type, would actually *prevent* TSM from working properly
under FreeBSD...
Cheers,
Paul.
--
e-mail: paul@gromit.dlib.vt.edu
"Without music to decorate it, time is just a bunch of boring production
deadlines or dates by which bills must be paid."
--- Frank Vincent Zappa
_______________________________________________
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscribe@freebsd.org"
- Previous message: Poul-Henning Kamp: "Re: TEST/REVIEW: minor improvement to linux-o-lator"
- In reply to: Thyer, Matthew: "RE: Q's about IBM TSM (was Re: HEADSUP: ibcs2 and svr4 comp at headed for history)"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|
|
