Re: Slug dies
--- In firstname.lastname@example.org, Nicola Fankhauser
> what is the file system of /share? if it's something other than ext3,The file system is ext3.
> reiserfs, xfs (or whatever linux really likes), then consider switching.
> is the dying process that causes the kernel oops always the sambaI'm not at all clear which process is causing the problem. I
> process? it could help to shut down samba during backups.
don't get enough information saved in my logs to tell. In this
most recent case, the only log with any useful into was the
rsnapshot log which showed first an "rsync" error, followed
by a "cp" error. Certainly I have seen a samba error in the
> Roger Shepherd wrote:Mmmm. Although I'm quite into the idea of playing around with
> > Any suggestions?
> you could try debianslug. this will be the more LSB conforming than your
> current setup seems to be (who on earth puts binaries other than
> black-box software like oracle etc. in /opt/bin? IMO they belong to
> /usr/local/bin). take this with a grain of salt, since I'm relatively
> new to the slug world and a big fan of debian since ages. however the
> ARM-debian tree is so complete, functional and working I cannot really
> understand people not wanting to work with it...
> different idea (or combine it with debianslug): you could try using
> rsync directly and set it up to do about the same as rsnapshot.
the Slug, I'm not really into the time it takes. I would have
been happy to stick with the standard firmware if I'd been able
to do the types of backups I'm doing now (main disk backed up to
the second disk hourly, creating snapshots). I'm inclinded to
think that a move to the new version of unslung and a reinstall
of rsnapshot might solve my problem (- I think it has a newer
version of samba? -) but I know that will be a weekend's work;
and it might not work.
I'll see how things go and if I can capture a half decent log
which gives more clue as to what has been going on. Oh well,
off to my basement to get my laptop to recheck the integrity
of the file system.
- --- In email@example.com, "Roger Shepherd" <rog@...> wrote:
> > is the dying process that causes the kernel oops always the sambaActually, rechecking the last set of logs, there is one entry in
> > process? it could help to shut down samba during backups.
log.smbd which might mean something:
[2006/06/04 00:25:43, 0] lib/util_sock.c:read_data(436)
read_data: read failure for 4. Error = No route to host
[2006/06/04 10:36:16, 0] smbd/server.c:main(708)
This occurred before the penultimate rsnapshot run (which
appears to have been OK).