[prev in list] [next in list] [prev in thread] [next in thread] 

List:       evms-devel
Subject:    [Evms-devel] Re: replace and expand/some problems
From:       Ross Boylan <RossBoylan () stanfordalumni ! org>
Date:       2004-05-19 5:06:19
Message-ID: 20040519050619.GA2067 () wheat ! boylan ! org
[Download RAW message or body]

After the moves described in my previous message (reproduced below) I
ran into some problems.  I'm frankly not sure any of them have much to
do with evms, but I thought I'd report them here.  I seem to have
recovered fully from the problems, except for the original one in my
earlier report.

On Tue, May 18, 2004 at 04:00:38PM -0700, Ross Boylan wrote:
> I put in a new disk, attached a DOS segment manager to it, and created
> some segments.  I made a new LVM container in one, and a 1G region in
> the container.
> 
> My old disk had an LVM container with a a 600MB region, used to make
> the volume /dev/evms/var.
> 
> I did a replace on the old region with the new region.  I got a
> message saying this had completed successfully.
> 
> df /var still shows a 600M disk.
> emvsgui shows a 600M logical volume sitting on top of a 1G region.
> When I try to expand the volume, I seem to end up with a dialog
> offering to expand the region
[snip]

After all the above, my system locked up badly.  The problem had
nothing to do with evms (unless the previous operations had some
mysterious long-range effect).  [I think my daughter did ctl-alt-F8 to
get to "window", found it blank (expected), and immediately did
ctl-alt-F7.  (Debian GNU/Linux running XFree 4.3.0 under KDE) and
somehow ended up blanking the screen and locking the keyboard.  The
keyboard showed green lights rotating on and off between caps lock and
something else I don't know (has a down arrow icon).]

I had to do a hardware reset, and got warnings of fatal corruption on
two Reiser volumes ("fatal corruptions, use --rebuild tree") during
reboot.  The first was in an old evms managed LVM container.  I did
nothing to it in my earlier manipulations, and in fact I doubt it had
much activity at all.

The second volume was the one I had just replaced, mounted as /var.  I
ended up with lots of files in lost+found (243) after the manual fsck.
I've checked a handful of the lost files, and all of them look like
either the output of lsmod or dumps of symbols or stack traces.  Dates
range back to May 3.

Finally, I noticed that whenever I ran fsck.reiserfs --rebuild tree I
got the message "unknown block ioctl 0x80081272".  I reported about a
week ago noticing some odd unknown ioctl errors on boot from
devmapper.  I don't recall if these were from device-mapper; I thought
they were, but that doesn't appear in my notes.

Running evms 2.3.1-1.

As far as I can tell, everything is OK on both volumes with the
reported fatal corruptions.

After the reboot, evms gui  shows /dev/evms/var (as I call the device
for my /var partition) as 1G.  However, details for that volume shows
a file system size of 600M.  df /var still shows it as 600M.

As before, I see no way with the gui to expand the actual 600M file
system to fill the region.


-------------------------------------------------------
This SF.Net email is sponsored by: SourceForge.net Broadband
Sign-up now for SourceForge Broadband and get the fastest
6.0/768 connection for only $19.95/mo for the first 3 months!
http://ads.osdn.com/?ad_id=2562&alloc_id=6184&op=click
_______________________________________________
Evms-devel mailing list
Evms-devel@lists.sourceforge.net
To subscribe/unsubscribe, please visit:
https://lists.sourceforge.net/lists/listinfo/evms-devel
[prev in list] [next in list] [prev in thread] [next in thread]