On Friday 31 March 2006 21:06, Warren Turkal wrote:
1. SIGSTOP all process with the pgrp of cyrmaster
2. ctl_cyrusdb -c
- optionally ctl_mboxlist -d here
Are the STOP and cyrusdb -c really necessary here? I believe the docs state
that only minor inconsistencies would occur with LVM
I know this is a perennial question, but I wanted to see if the following
strategy would work for a hot backup for Cyrus.
1. SIGSTOP all process with the pgrp of cyrmaster
2. ctl_cyrusdb -c
- optionally ctl_mboxlist -d here
3. create lvm ro snapshot
4. SIGCONT the processes with pgrp of
Have a few questions about back ups, well, just one.
Lets say my site required to be up 24/7... so the only way we could do
backups of cyrus would be live.
Certainly one can just copy the entire mail spool... getting MOST of hte
messages, and the .headers, .cache, .index, etc. Is this safe?
I'm
Certainly one can just copy the entire mail spool... getting MOST of hte
messages, and the .headers, .cache, .index, etc. Is this safe?
Personally, I wouldn't just copy the spool from beginning to end while
cyrus was running - all your indexes etc will become out of sync if mail
is delivered to
I'm also running into an issue where occasionally the snapshot fails -
it's taken while reiserfs is in the middle of a transaction, and the
journal can't be replayed on mount because the snapshot is read-only.
This is more of an LVM issue than anything else. I get around it by
syncing the
I've been using XFS on LVM and I think it doesn't have the problem you
describe. With XFS, you have to call 'xfs_freeze -f' before taking the
snapshot, then you 'xfs_freeze -u' it again.
That's exactly what I need for reiserfs :-( . Darn. I guess that's
something to note in the Cyrus FAQ - xfs
On Fri, 14 Nov 2003, Simon Matter wrote:
Would be nice to have something like a 'cyrus_freeze' command. Maybe there
is a way to do this with SIGSTOP/SIGCONT?
At this point, you might as well just kill the server, since you won't be
able to do almost anything for the duration of the snapshot.
On Fri, 14 Nov 2003, Jerry Haltom wrote:
But if one can just reconstruct the indexes, what damnage can be done by
just copying them? What critical information do they hold?
If the index file goes missing, you'll lose most flag information (not
seen state). Reconstruct tries to use this data
But if one can just reconstruct the indexes, what damnage can be done by
just copying them? What critical information do they hold?
On Fri, 2003-11-14 at 08:54, Rob Siemborski wrote:
On Fri, 14 Nov 2003, Simon Matter wrote:
Would be nice to have something like a 'cyrus_freeze' command.
On Fri, 14 Nov 2003, Simon Matter wrote:
IIRC taking a snapshot with LVM/XFS takes just a few seconds. I guess
everything should continue after the snapshot. What do the cyrus processes
do if they receive SIGSTOP? Will they stop immediately or will they not
stop at all? I mean, if a process
On Fri, 14 Nov 2003, Simon Matter wrote:
Would be nice to have something like a 'cyrus_freeze' command. Maybe
there
is a way to do this with SIGSTOP/SIGCONT?
At this point, you might as well just kill the server, since you won't be
able to do almost anything for the duration of the
On Fri, 14 Nov 2003, Rob Siemborski wrote:
On Fri, 14 Nov 2003, Simon Matter wrote:
IIRC taking a snapshot with LVM/XFS takes just a few seconds. I guess
everything should continue after the snapshot. What do the cyrus processes
do if they receive SIGSTOP? Will they stop immediately or
On Fri, 14 Nov 2003, Rob Siemborski wrote:
On Fri, 14 Nov 2003, Simon Matter wrote:
IIRC taking a snapshot with LVM/XFS takes just a few seconds. I guess
everything should continue after the snapshot. What do the cyrus
processes
do if they receive SIGSTOP? Will they stop immediately
Actually I do have a perl script that I wrote that recreates the
mailboxes.db file if it is lost.. I had this happen when I was upgrading
from 2.0 to 2.1 on my system (my personal mail box only)..
The script only does the user directory and just adds in all the folders
with the permissions of
On Fri, 14 Nov 2003, James Miller wrote:
Rob, I'm still abit confused about what's needed to get cyrus (including
index files) into a consistent state -- ready for a backup/snapshot.. etc.
etc.
Right now? Stop it, snapshot the volume, start it again. It should take a
few seconds, which is an
15 matches
Mail list logo