Coda File System

Re: Backup problems

From: Jan Harkes <jaharkes_at_cs.cmu.edu>
Date: Wed, 13 Jun 2001 07:59:46 -0400
On Wed, Jun 13, 2001 at 12:00:06PM +0200, Andrea Cerrito wrote:
> Hi to all.
> I'm trying to backup my volumes, and I set /vice/db/dumplist and
> /vice/db/vicetab for my needs.

First off, Coda's own backup tool first dumps all volume to the holding
area(s) and then spools them to tape. However, it looks like you have
enough server diskspace to store up to 84GB of data, but the backup
partition is about 8GB.

In this case, using Amanda for the backups would actually be benificial,
Amanda schedules the backups in such a way that the backup load is
spread across the dumpcycle, and it will interleave pulling the data
from the Coda servers with writing to tape.

> As you can see it's empty.
> But when I try to start the backup:
> =============================
> [root_at_apu-a /root]# backup -t 135 /vice/db/dumplist /extra0

Ah, I see the problem.

The directory you specify at the end of the backup command line should
be a different name. It will be filled with symlinks to the actual
volumes in the backup partitions that are configured using vicetab.
(Don't ask me why, it seems a bit redundant to me).

So create an empty directory f.i. /backup that is not listed in vicetab,
and run the backup as follows,

    backup -t 135 /vice/db/dumplist /backup 

> What I have to do?
> Restart all servers? I don't want to stop them (it's a production machine):
> is there a way to do a backup on a volume by hand?

There should be no need to restart the servers. Yes, a volume can be
backed up by hand using the following volutil commands.

    volutil -h <servername> lock <volume-replica-id>
    volutil -h <servername> backup <volume-replica-id>
    volutil -h <servername> dump <backup-volume-id> [dumpfile]

Where volume replica id is not the replicated group id that is listed in
the dumplist file. Similarily the backup command creates a read-only
snapshot of the volume state which is assigned it's own unique volume-id
that is used to dump the volume. Typically I get these using 'volutil info'

volutil -h srv info vm:u.jaharkes.mail | grep '^id ='
id = d1000018, parentId = d1000018, cloneId = 0, backupId = d1000099, ...
     ^^^^^^^^ volume-replica id            backup volume id ^^^^^^^^
					   (only valid after volutil
					   backup is run)

Jan
Received on 2001-06-13 07:59:50