On Montag, 21. Juni 2004 18:40, Martin Michlmayr wrote:
* Matthias Murra [EMAIL PROTECTED] [2004-06-21 17:17]:
It's certainly not lvm's fault if the installer screwed up a bit,
so yes, this bug probably can be closed -- or reassigned to the
So what exactly did the installer do? From what I
I've not had a play with the installer for some time (to my shame) but I wonder
if it's at all possible to notice that a disk that was allocated to LVM
(pvcreated) has now been allocated to a filesystem. In which case a pvremove
could be done on the volume.
This maybe totally impractical, of
* Patrick Caulfield [EMAIL PROTECTED] [2004-06-22 09:14]:
I've not had a play with the installer for some time (to my shame) but I wonder
if it's at all possible to notice that a disk that was allocated to LVM
(pvcreated) has now been allocated to a filesystem. In which case a pvremove
could
On Tue, Jun 22, 2004 at 01:46:53PM +0100, Martin Michlmayr wrote:
* Patrick Caulfield [EMAIL PROTECTED] [2004-06-22 09:14]:
I've not had a play with the installer for some time (to my shame) but I wonder
if it's at all possible to notice that a disk that was allocated to LVM
(pvcreated) has
On Sun, Jun 20, 2004 at 07:33:45PM +0200, Matthias Murra wrote:
On Freitag, 18. Juni 2004 18:12, Patrick Caulfield wrote:
On Fri, Jun 18, 2004 at 05:54:38PM +0200, Matthias Murra wrote:
You probably shouldn't need to reinstall. Upgrading the tools
should either fix it automatically, or
On Montag, 21. Juni 2004 08:22, Patrick Caulfield wrote:
Well, if things seem to be working OK, I wouldn't worry too much
about it! I would have expected vgchange to fail if the MDA
checksums were wrong but it seems not.
In the meantime can you send me a vgdisplay output and a
pvdisplay for
Hang on, there's an alarm bell that I've not been heeding.
Looking in the code: Incorrect metadata area header checksum is a fatal error.
If that's is occurring on one of tour PVs then the VG should not activate.that
leads me to think it might be another disk/device that is producing the error.
On Montag, 21. Juni 2004 14:24, Patrick Caulfield wrote:
Hang on, there's an alarm bell that I've not been heeding.
Looking in the code: Incorrect metadata area header checksum is a
fatal error.
If that's is occurring on one of tour PVs then the VG should not
activate.that leads me to think
On Mon, Jun 21, 2004 at 03:58:37PM +0200, Matthias Murra wrote:
On Montag, 21. Juni 2004 14:24, Patrick Caulfield wrote:
Hang on, there's an alarm bell that I've not been heeding.
Looking in the code: Incorrect metadata area header checksum is a
fatal error.
If that's is occurring on
On Montag, 21. Juni 2004 16:18, Patrick Caulfield wrote:
Ah, so you did pvcreate on it (or the installer did), then did mkfs
over that? I can imagine that would have an odd effect on the LVM
metadata!
The installer seems to have done that, yes. I thought the installer's
behavior was pretty
* Matthias Murra [EMAIL PROTECTED] [2004-06-21 17:17]:
It's certainly not lvm's fault if the installer screwed up a bit, so
yes, this bug probably can be closed -- or reassigned to the
So what exactly did the installer do? From what I can see, you
created some LVM devices, but they were
On Freitag, 18. Juni 2004 18:12, Patrick Caulfield wrote:
On Fri, Jun 18, 2004 at 05:54:38PM +0200, Matthias Murra wrote:
You probably shouldn't need to reinstall. Upgrading the tools
should either fix it automatically, or you can do
vgcfgbackup/vgcfgrestore to rewrite the metadata.
I have
On Fri, 11 Jun, 2004, at 17:25:25 +0100, Patrick Caulfield wrote:
On Fri, Jun 11, 2004 at 03:10:18PM +0100, Martin Michlmayr wrote:
* [EMAIL PROTECTED] [EMAIL PROTECTED] [2004-06-11 06:19]:
In any case, can you type:
vgchange -a y
and restart the partitioner; then they should show
On Fri, Jun 18, 2004 at 05:54:38PM +0200, Matthias Murra wrote:
I have stumbled across this thread by googling for the incorrect
metadata area header checksum message mentioned above, because
that's the message I am seeing when booting my Debian Sid system that
I installed on my Toshiba
On Sun, Jun 13, 2004 at 02:43:29PM +0100, Martin Michlmayr wrote:
* Patrick Caulfield [EMAIL PROTECTED] [2004-06-11 17:25]:
There was an odd bug in earlier lvm2 that used to cause checksum
failures but I can't locate which version that was and I assume that
you'll be using a fairly recent
* Patrick Caulfield [EMAIL PROTECTED] [2004-06-14 09:17]:
http://people.debian.org/~patrick/lvm2-udeb_2.00.16-2_i386.udeb
daveg2: After network configuration in debian-installer, open a shell
on the 2nd virtual console (alt-f2) and then type:
wget
* Patrick Caulfield [EMAIL PROTECTED] [2004-06-11 17:25]:
There was an odd bug in earlier lvm2 that used to cause checksum
failures but I can't locate which version that was and I assume that
you'll be using a fairly recent udeb anyway.
No, the udeb we currently use is 2.00.08-4 (at least on
* [EMAIL PROTECTED] [EMAIL PROTECTED] [2004-06-10 19:42]:
lvscan reports the same first three lines as vgscan and concludes
by correctly listing all lvs, but as inactive. Here's the first listed:
inactive '/dev/vg1/usrlv' [1.0 GB] next free (default).
Strange, I wonder why they're
On 11 Jun 2004 at 13:13, Martin Michlmayr wrote:
Strange, I wonder why they're inactive... since you just created them,
they should be active. You did just create them, right? Or did they
exist before?
I did just create them. In other runs where I did not, it tells me that
it found
reassign 252164 lvm2
thanks
* [EMAIL PROTECTED] [EMAIL PROTECTED] [2004-06-11 06:19]:
In any case, can you type:
vgchange -a y
and restart the partitioner; then they should show up.
Done. For each lv, I get:
Incorrect metadata area header checksum
Incorrect metadata area header
On Fri, Jun 11, 2004 at 03:10:18PM +0100, Martin Michlmayr wrote:
reassign 252164 lvm2
thanks
* [EMAIL PROTECTED] [EMAIL PROTECTED] [2004-06-11 06:19]:
In any case, can you type:
vgchange -a y
and restart the partitioner; then they should show up.
Done. For each lv, I get:
* [EMAIL PROTECTED] [EMAIL PROTECTED] [2004-06-02 22:21]:
OK, I went back through the steps. Indeed I did all of 1 through 5 in
the correct order. I can't for the life of me figure out 6. In the main
menu of Partition Disks, I can see all partitions including the two
lvm partitions, but
On 10 Jun 2004 at 20:48, Martin Michlmayr wrote:
* [EMAIL PROTECTED] [EMAIL PROTECTED] [2004-06-02 22:21]:
OK, I went back through the steps. Indeed I did all of 1 through 5 in
the correct order. I can't for the life of me figure out 6. In the main
menu of Partition Disks, I can see
On Wed, Jun 02, 2004 at 09:11:11PM -0700, [EMAIL PROTECTED] wrote:
I just today discovered the IBM donation of EVMS to the LINUX world. I'll bet I
also
need to select evms-udeb? I have not been because I didn't know what EVMS is. If
not required, what does evms-udeb add?
evms-udeb is
On 1 Jun 2004 at 19:50, Margarita Manterola wrote:
Hola daveg!
Partitioning goes fine, configuring the lvm goes fine,
but on leaving the lvm and starting up the partitioner,
it hangs at 56%. Only once out of many attempts did
it complete. In that case, it continued to install base
On 1 Jun 2004 at 19:50, Margarita Manterola wrote:
Hola daveg!
Well, this might seem stupid, but did you select the filesystem in the
partition manager menu? That's where you have to select it. These are
the steps for LVM in the new debian installer:
1 - Create the LVM partitions in the
Package: installation-reports
Debian-installer-version:
Sarge daily 5-22-04 from the Debian site
uname -a:
linux server 2.6.5-1-386#2 Fri Apr 30 20:13:30 EST 2004 i586 unknown
Date: Friday May 7, 2004 ~8:30 PM PDT
Method: Net install using the 800 MB netinst image,
booted from the CD,
* [EMAIL PROTECTED] [EMAIL PROTECTED] [2004-06-01 12:45]:
Sarge daily 5-22-04 from the Debian site
Can you get a newer image from
http://gluck.debian.org/cdimage/testing/sarge_d-i/
(although it shouldn't really matter).
Partitioning goes fine, configuring the lvm goes fine, but on
leaving the
Hola [EMAIL PROTECTED]
Partitioning goes fine, configuring the lvm goes fine,
but on leaving the lvm and starting up the partitioner,
it hangs at 56%. Only once out of many attempts did
it complete. In that case, it continued to install base
system without offering to create
29 matches
Mail list logo
