[linux-lvm] Can no longer mount lvm striped volume

[linux-lvm] Can no longer mount lvm striped volume

[Craig Tierney]

I just had a power outage and now I cannot
remount my LVM volumes.  I have to volume
groups each with one logical volume.  Each
vg is striped across two luns.

I started getting the errors:

vgscan -- ERROR: logical volume "/dev/sdd_vol2/lv1" reuses 
an existing device number; please vgexport/vgimport that VG 

I was using lvm-1.0.2 at the time so I could
not run vgscan -f.  So I upgraded my 2.4.14 kernel
to lvm-1.1-rc2 and gave it a try.

Now I vgscan will not complain about the logical
volumes, but I still cannot mount the ext3 filesystems
that are installed on the volumes.

I am now getting the error message:

[root@x06 /root]# mount -t ext3 /dev/sdd_vol2/ /export/sdd_vol2
mount: /dev/sdd_vol2/ is not a block device

IO nodes have gone down in the past due to
power interruptions, but I have never had
this catastrophic of a problem.  I have 16 volume
groups, and 3 or 4 appear to be affected.

Any ideas?

Thanks,
Craig


-- 
Craig Tierney (ctierney@hpti.com)

Re: [linux-lvm] Can no longer mount lvm striped volume

[John Moser]

> [root@x06 /root]# mount -t ext3 /dev/sdd_vol2/ /export/sdd_vol2
> mount: /dev/sdd_vol2/ is not a block device
>
> IO nodes have gone down in the past due to
> power interruptions, but I have never had
> this catastrophic of a problem.  I have 16 volume
> groups, and 3 or 4 appear to be affected.

Have you tried a simple:

vgchange -a y

?  Similar situation happened to me before, and I just needed to activate
the volume group.

-John

Re: [linux-lvm] Can no longer mount lvm striped volume

[Craig Tierney]

As Homer would say 'Doh!'.

I am still having some other problems, but this error
was fixed by mounting the logical volume and not the
volume group.

Wrong:  mount -t ext3 /dev/sdd_vol2 /export/sdd_vol2
Correct: mount -t ext3 /dev/sdd_vol2/lv1 /export/sdd_vol2

I was still having problems with the scan message, but
after I fixed that I didn't run the mounting scripts I
mounted by hand and messed up.

Craig


On Thu, May 02, 2002 at 01:34:40PM -0500, John Moser wrote:
> 
> > [root@x06 /root]# mount -t ext3 /dev/sdd_vol2/ /export/sdd_vol2
> > mount: /dev/sdd_vol2/ is not a block device
> >
> > IO nodes have gone down in the past due to
> > power interruptions, but I have never had
> > this catastrophic of a problem.  I have 16 volume
> > groups, and 3 or 4 appear to be affected.
> 
> Have you tried a simple:
> 
> vgchange -a y
> 
> ?  Similar situation happened to me before, and I just needed to activate
> the volume group.
> 
> -John
> 
> 
> _______________________________________________
> linux-lvm mailing list
> linux-lvm@sistina.com
> http://lists.sistina.com/mailman/listinfo/linux-lvm
> read the LVM HOW-TO at http://www.sistina.com/lvm/Pages/howto.html

-- 
Craig Tierney (ctierney@hpti.com)

Re: [linux-lvm] Can no longer mount lvm striped volume

[Craig Tierney]

Things are getting better.  I found 2 of the
3 lv's that were missing.  I am trying to find the last
one.  

All of the vg's and lv's were created under lvm 1.0.2.
I am now using 1.1-rc2.

If I type either:

   vgdisplay -v sdd_vol4

or 

   vgdisplay -v sdd2_vol12

I get the same results, both for sdd2_vol12.
That vg does not yet have a lv.  I get the same
results from different nodes (storage is fiber attached).

Is there a maximum number of vg's that are supported?
Is there a maximum length that a vg name can be?
  
When I first configured the new disks I had some problems
where trying to run vgdisplay on the volumes named 'sdd_vol1?' 
I would only get the results for sdd_vol1.  Is there an
8 character max built in?

Craig

On Thu, May 02, 2002 at 01:34:40PM -0500, John Moser wrote:
> 
> > [root@x06 /root]# mount -t ext3 /dev/sdd_vol2/ /export/sdd_vol2
> > mount: /dev/sdd_vol2/ is not a block device
> >
> > IO nodes have gone down in the past due to
> > power interruptions, but I have never had
> > this catastrophic of a problem.  I have 16 volume
> > groups, and 3 or 4 appear to be affected.
> 
> Have you tried a simple:
> 
> vgchange -a y
> 
> ?  Similar situation happened to me before, and I just needed to activate
> the volume group.
> 
> -John
> 
> 
> _______________________________________________
> linux-lvm mailing list
> linux-lvm@sistina.com
> http://lists.sistina.com/mailman/listinfo/linux-lvm
> read the LVM HOW-TO at http://www.sistina.com/lvm/Pages/howto.html

-- 
Craig Tierney (ctierney@hpti.com)

Re: [linux-lvm] Can no longer mount lvm striped volume

[Patrick Caulfield]

On Thu, May 02, 2002 at 01:07:22PM -0600, Craig Tierney wrote:
> Things are getting better.  I found 2 of the
> 3 lv's that were missing.  I am trying to find the last
> one.  

Try vgscan -f, it will replace all the LV numbers with new ones rather than try
to preserve them.

-- 

patrick

Re: [linux-lvm] Can no longer mount lvm striped volume

[Craig Tierney]

On Thu, May 02, 2002 at 08:36:10PM +0100, Patrick Caulfield wrote:
> On Thu, May 02, 2002 at 01:07:22PM -0600, Craig Tierney wrote:
> > Things are getting better.  I found 2 of the
> > 3 lv's that were missing.  I am trying to find the last
> > one.  
> 
> Try vgscan -f, it will replace all the LV numbers with new ones rather than try
> to preserve them.

I tried vgscan -f, but it did not help.

I get different results if I run vgdisplay with and without
the -D option.  With the -D option, the volume group looks correct.
I did remember to update the lvm tools and not just the kernel
patch.

(The wrong results)
[root@x05 /root]# vgdisplay -v sdd_vol4
--- Volume group ---
VG Name               sdd2_vol12 
VG Access             read/write
VG Status             available/resizable
VG #                  19
MAX LV                255
Cur LV                0
Open LV               0
MAX LV Size           511.98 GB
Max PV                255
Cur PV                1
Act PV                1
VG Size               499.98 GB
PE Size               8 MB
Total PE              63998
Alloc PE / Size       0 / 0
Free  PE / Size       63998 / 499.98 GB
VG UUID               f9oa9Y-OsTL-Yypp-2KCK-kVzG-Vinb-c2eQlg

--- No logical volumes in "sdd2_vol12" ---


--- Physical volumes ---
PV Name (#)           /dev/sdaa (1)
PV Status             available / allocatable
Total PE / Free PE    63998 / 63998


(The right results)
[root@x05 /root]# vgdisplay -D -v sdd_vol4
--- Volume group ---
VG Name               sdd_vol4 
VG Access             read/write
VG Status             NOT available/resizable
VG #                  19
MAX LV                255
Cur LV                1
Open LV               0
MAX LV Size           511.98 GB
Max PV                255
Cur PV                2
Act PV                2
VG Size               488.27 GB
PE Size               8 MB
Total PE              62498
Alloc PE / Size       62498 / 488.27 GB
Free  PE / Size       0 / 0
VG UUID               cBSzp4-kauE-Tv0N-SC3r-6mAi-RGw0-zwTbDC

--- Logical volume ---
LV Name                /dev/sdd_vol4/lv1
VG Name                sdd_vol4
LV Write Access        read/write
LV Status              available
LV #                   1
# open                 0
LV Size                488.27 GB
Current LE             62498
Allocated LE           62498
Stripes                2
Stripe size (KByte)    64
Allocation             next free
Read ahead sectors     10000
Block device           58:7


--- Physical volumes ---
PV Name (#)           /dev/sdo (1)
PV Status             NOT available / allocatable
Total PE / Free PE    31249 / 0

PV Name (#)           /dev/sdh (2)
PV Status             NOT available / allocatable
Total PE / Free PE    31249 / 0

Re: [linux-lvm] Can no longer mount lvm striped volume

[Craig Tierney]

Problem solved.  I rebooted the node, and I still had problems.
Then I explicitly ran "vgchange -a y sdd_vol4".  That seemed to
do the trick.  My startup scripts do run "vgchange -a y" which
wasn't adequate.

Thanks for the help.
Craig



> On Thu, May 02, 2002 at 08:36:10PM +0100, Patrick Caulfield wrote:
> > On Thu, May 02, 2002 at 01:07:22PM -0600, Craig Tierney wrote:
> > > Things are getting better.  I found 2 of the
> > > 3 lv's that were missing.  I am trying to find the last
> > > one.  
> > 
> > Try vgscan -f, it will replace all the LV numbers with new ones rather than try
> > to preserve them.
> 
> I tried vgscan -f, but it did not help.
> 
> I get different results if I run vgdisplay with and without
> the -D option.  With the -D option, the volume group looks correct.
> I did remember to update the lvm tools and not just the kernel
> patch.
> 
> (The wrong results)
> [root@x05 /root]# vgdisplay -v sdd_vol4
> --- Volume group ---
> VG Name               sdd2_vol12 
> VG Access             read/write
> VG Status             available/resizable
> VG #                  19
> MAX LV                255
> Cur LV                0
> Open LV               0
> MAX LV Size           511.98 GB
> Max PV                255
> Cur PV                1
> Act PV                1
> VG Size               499.98 GB
> PE Size               8 MB
> Total PE              63998
> Alloc PE / Size       0 / 0
> Free  PE / Size       63998 / 499.98 GB
> VG UUID               f9oa9Y-OsTL-Yypp-2KCK-kVzG-Vinb-c2eQlg
> 
> --- No logical volumes in "sdd2_vol12" ---
> 
> 
> --- Physical volumes ---
> PV Name (#)           /dev/sdaa (1)
> PV Status             available / allocatable
> Total PE / Free PE    63998 / 63998
> 
> 
> (The right results)
> [root@x05 /root]# vgdisplay -D -v sdd_vol4
> --- Volume group ---
> VG Name               sdd_vol4 
> VG Access             read/write
> VG Status             NOT available/resizable
> VG #                  19
> MAX LV                255
> Cur LV                1
> Open LV               0
> MAX LV Size           511.98 GB
> Max PV                255
> Cur PV                2
> Act PV                2
> VG Size               488.27 GB
> PE Size               8 MB
> Total PE              62498
> Alloc PE / Size       62498 / 488.27 GB
> Free  PE / Size       0 / 0
> VG UUID               cBSzp4-kauE-Tv0N-SC3r-6mAi-RGw0-zwTbDC
> 
> --- Logical volume ---
> LV Name                /dev/sdd_vol4/lv1
> VG Name                sdd_vol4
> LV Write Access        read/write
> LV Status              available
> LV #                   1
> # open                 0
> LV Size                488.27 GB
> Current LE             62498
> Allocated LE           62498
> Stripes                2
> Stripe size (KByte)    64
> Allocation             next free
> Read ahead sectors     10000
> Block device           58:7
> 
> 
> --- Physical volumes ---
> PV Name (#)           /dev/sdo (1)
> PV Status             NOT available / allocatable
> Total PE / Free PE    31249 / 0
> 
> PV Name (#)           /dev/sdh (2)
> PV Status             NOT available / allocatable
> Total PE / Free PE    31249 / 0
> 
> 
> 
> 
> _______________________________________________
> linux-lvm mailing list
> linux-lvm@sistina.com
> http://lists.sistina.com/mailman/listinfo/linux-lvm
> read the LVM HOW-TO at http://www.sistina.com/lvm/Pages/howto.html

-- 
Craig Tierney (ctierney@hpti.com)

Re: [linux-lvm] Can no longer mount lvm striped volume

[Heinz J . Mauelshagen]

Craig,

you said, "vgscan -f" didn't enable you to "vgchange -ay" the VG in question
but the explicit "vgchange -ay VGName" helped?

Very strange indeed.

Could you send the output of "dd if=$PV of=$(basename $PV).vgda bs=1k count=4k"
(replace the PV variable by your PV device pathes in turn) plus /etc/lvmtab*
contained in a bzip2ed tar archive to me directly <mge@sistina.com> for
further analisys.

Thanks,
Heinz    -- The LVM Guy --

On Thu, May 02, 2002 at 12:48:50PM -0600, Craig Tierney wrote:
> As Homer would say 'Doh!'.
> 
> I am still having some other problems, but this error
> was fixed by mounting the logical volume and not the
> volume group.
> 
> Wrong:  mount -t ext3 /dev/sdd_vol2 /export/sdd_vol2
> Correct: mount -t ext3 /dev/sdd_vol2/lv1 /export/sdd_vol2
> 
> I was still having problems with the scan message, but
> after I fixed that I didn't run the mounting scripts I
> mounted by hand and messed up.
> 
> Craig
> 
> 
> On Thu, May 02, 2002 at 01:34:40PM -0500, John Moser wrote:
> > 
> > > [root@x06 /root]# mount -t ext3 /dev/sdd_vol2/ /export/sdd_vol2
> > > mount: /dev/sdd_vol2/ is not a block device
> > >
> > > IO nodes have gone down in the past due to
> > > power interruptions, but I have never had
> > > this catastrophic of a problem.  I have 16 volume
> > > groups, and 3 or 4 appear to be affected.
> > 
> > Have you tried a simple:
> > 
> > vgchange -a y
> > 
> > ?  Similar situation happened to me before, and I just needed to activate
> > the volume group.
> > 
> > -John
> > 
> > 
> > _______________________________________________
> > linux-lvm mailing list
> > linux-lvm@sistina.com
> > http://lists.sistina.com/mailman/listinfo/linux-lvm
> > read the LVM HOW-TO at http://www.sistina.com/lvm/Pages/howto.html
> 
> -- 
> Craig Tierney (ctierney@hpti.com)
> 
> _______________________________________________
> linux-lvm mailing list
> linux-lvm@sistina.com
> http://lists.sistina.com/mailman/listinfo/linux-lvm
> read the LVM HOW-TO at http://www.sistina.com/lvm/Pages/howto.html

*** Software bugs are stupid.
    Nevertheless it needs not so stupid people to solve them ***

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Heinz Mauelshagen                                 Sistina Software Inc.
Senior Consultant/Developer                       Am Sonnenhang 11
                                                  56242 Marienrachdorf
                                                  Germany
Mauelshagen@Sistina.com                           +49 2626 141200
                                                       FAX 924446
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-