[linux-lvm] linux-lvm Digest, Vol 70, Issue 6

Johan Gardell gardin at gmail.com
Sun Dec 20 13:36:21 UTC 2009


Hi!

I was able to use dd to read from the two PVs, so i started looking in
the /proc/* files as you said:

/proc/mounts:
rootfs / rootfs rw 0 0
none /sys sysfs rw,nosuid,nodev,noexec,relatime 0 0
none /proc proc rw,nosuid,nodev,noexec,relatime 0 0
udev /dev tmpfs rw,relatime,size=10240k,mode=755 0 0
/dev/disk/by-uuid/56b77d5f-64e8-4392-be98-81a17a4c895d / reiserfs
rw,relatime,notail 0 0
tmpfs /lib/init/rw tmpfs rw,nosuid,relatime,mode=755 0 0
tmpfs /dev/shm tmpfs rw,nosuid,nodev,relatime 0 0
devpts /dev/pts devpts rw,nosuid,noexec,relatime,gid=5,mode=620,ptmxmode=000 0 0
/dev/mapper/Dreamhack-dreamhacklv /home reiserfs rw,relatime 0 0

and in /proc/swap
Filename				Type		Size	Used	Priority
/dev/sdb3                               partition	1951888	0	-1

Which did not give much, but i looked in /proc/mdstat as well
Personalities :
md127 : inactive sdc[1](S) sda[0](S)
      4514 blocks super external:imsm

unused devices: <none>

Doesn't that mean the PVs are mapped to the md127 device? (the
/dev/md127 device exists as well)
I googled some and found that mdadm --remove /dev/md127 /dev/sdc and
the same for the other device should remove the devices from the raid,
but i just want to check with you so that i don't toast my data.. :P
Could this be the issue?

Thanks in regards :)
Johan Gardell

2009/12/18  <linux-lvm-request at redhat.com>:
>> Message: 3
>> Date: Thu, 10 Dec 2009 10:40:53 -0800
>> From: malahal at us.ibm.com
>> Subject: Re: [linux-lvm] Re: Re: Problems with dissapearing PV when
>>        mounting (Stuart D. Gathman)
>> To: linux-lvm at redhat.com
>> Message-ID: <20091210184053.GA6645 at us.ibm.com>
>> Content-Type: text/plain; charset=us-ascii
>>
>> Johan Gardell [gardin at gmail.com] wrote:
>>> The output from
>>> dmsetup table
>>>   gardin-swap_1:
>>>   gardin-root:
>>>   Dreamhack-dreamhacklv: 0 2636726272 linear 8:34 384
>>
>> The above output clearly shows that gardin-root and gardin-swap_1 have
>> empty tables. I remember there was a fix that cleaned up empty tables
>> like above but that won't help you. You need to figure out the reason
>> for 'device mapper lookup failure' messages.
>>
>>> >> [22723.980390] device-mapper: table: 254:1: linear: dm-linear: Device
>>> >> lookup failed
>>> >> [22723.980395] device-mapper: ioctl: error adding target to table
>>> >> [22724.001153] device-mapper: table: 254:2: linear: dm-linear: Device
>>> >> lookup failed
>>> >> [22724.001158] device-mapper: ioctl: error adding target to table
>>
>> The above messages are direct reason for not "really" creating
>> gardin-root or gardin-swap_1 logical volumes.  If you know the
>> underlying PV's that make up gardin-root LV, try reading from them using
>> 'dd'. If you can read, then the most likely problem is that someone else
>> has opened them. It may be hard to find though. You need to go through
>> usual culprits by running, 'cat /proc/mounts', 'cat /proc/swaps' etc. Is
>> EVMS using them by any chance..., maybe mdraid...
>>
>>> >> mount doesn't give any messages in dmesg
>>> >>
>>> >> lvs shows:
>>> >> ? LV ? ? ? ? ?VG ? ? ? ?Attr ? LSize ? Origin Snap% ?Move Log Copy% ?Convert
>>> >> ? dreamhacklv Dreamhack -wi-ao ? 1,23t
>>> >> ? root ? ? ? ?gardin ? ?-wi-d- 928,00g
>>> >> ? swap_1 ? ? ?gardin ? ?-wi-d- ? 2,59g
>>> >>
>>> >> if i try to mount with:
>>> >> ? mount -t reiserfs /dev/mapper/gardin-root /mnt/tmp
>>> >>
>>> >> i get this in dmesg:
>>> >> ? [23113.711247] REISERFS warning (device dm-1): sh-2006
>>> >> read_super_block: bread failed (dev dm-1, block 2, size 4096)
>>> >> ? [23113.711257] REISERFS warning (device dm-1): sh-2006
>>> >> read_super_block: bread failed (dev dm-1, block 16, size 4096)
>>> >> ? [23113.711261] REISERFS warning (device dm-1): sh-2021
>>> >> reiserfs_fill_super: can not find reiserfs on dm-1
>>> >
>>> > Looks like you have some kind of LV here. What is the output of the
>>> > following two commands:
>>
>> Your 'dmsetup table' output shows why the mount failed. You have an LV
>> created with no mapping.
>>
>> Thanks, Malahal.
>>
>>
>>
>
>
>
> ------------------------------
>
> Message: 2
> Date: Fri, 11 Dec 2009 16:55:40 -0800
> From: malahal at us.ibm.com
> Subject: Re: [linux-lvm] Re: Re: Re: Problems with dissapearing PV
>        when    mounting (Stuart D. Gathman)
> To: linux-lvm at redhat.com
> Message-ID: <20091212005540.GA2562 at us.ibm.com>
> Content-Type: text/plain; charset=us-ascii
>
> Johan Gardell [gardin at gmail.com] wrote:
>> Ok, but why is that? I haven't changed anything on the LV. The output of
>> lvdisplay -m shows that i have physical volumes extents right? (but
>> /dev/gardin does not exist at all)
>
> "lvdisplay -m" shows what is in LVM metadata. The LVs just fail to
> activate in your case (root and swap_1). Since the LVM commands
> (lvdisplay) are able to read the metadata without any errors, I would
> assume that *some other software* on your machine openede the needed PVs
> before LVM did, so LVM is unable to activate the LVs (aka unable to load
> the mapping tables).
>
> Thanks, Malahal.
>
>
>
> ------------------------------
>
> Message: 3
> Date: Mon, 14 Dec 2009 14:45:21 +0100 (CET)
> From: Thomas Voegtle <tv at lio96.de>
> Subject: [linux-lvm] lvcreate does not respect units?
> To: linux-lvm at redhat.com
> Message-ID: <alpine.LNX.2.00.0912141234240.14618 at er-systems.de>
> Content-Type: TEXT/PLAIN; format=flowed; charset=US-ASCII
>
>
>
> Hi,
>
>
> I ran in a problem while creating a lv with units in powers of 1000.
>
> I looked up the free space with vgdisplay:
>
> root@(none):~# vgdisplay --units G | grep Size
> vgdisplay  VG Size               25.70 GB
> vgdisplay  PE Size               0.00 GB
> vgdisplay  Alloc PE / Size       0 / 0 GB
> vgdisplay  Free  PE / Size       6127 / 25.70 GB
>
>
> And then I tried to create the lv a little bit smaller, like this:
>
> root@(none):~# lvcreate -L 25.50G datavolume -n datavolume
> lvcreate  Insufficient free extents (6127) in volume group datavolume:
> 6528 required
>
>
> You even get the same message using a small g:
>
>
> root@(none):~# lvcreate -L 25.50g datavolume -n datavolume
> lvcreate  Insufficient free extents (6127) in volume group datavolume:
> 6528 required
>
>
> So it seems lvcreate does not respect the unit in power of 1000?
>
> I used 2.02.56.
>
>
>
> ------------------------------
>
> Message: 4
> Date: Mon, 14 Dec 2009 14:00:58 +0000
> From: Alasdair G Kergon <agk at redhat.com>
> Subject: Re: [linux-lvm] lvcreate does not respect units?
> To: LVM general discussion and development <linux-lvm at redhat.com>
> Message-ID: <20091214140057.GA14009 at agk-dp.fab.redhat.com>
> Content-Type: text/plain; charset=us-ascii
>
> On Mon, Dec 14, 2009 at 02:45:21PM +0100, Thomas Voegtle wrote:
>> vgdisplay  Free  PE / Size       6127 / 25.70 GB
>
>> root@(none):~# lvcreate -L 25.50g datavolume -n datavolume
>> lvcreate  Insufficient free extents (6127) in volume group datavolume:
>> 6528 required
>
>> So it seems lvcreate does not respect the unit in power of 1000?
>
> Use either:
>   lvcreate -l 6127
> or
>   lvcreate -l 100%FREE
>
> Alasdair
>
>
>
> ------------------------------
>
> Message: 5
> Date: Mon, 14 Dec 2009 15:28:24 +0100 (CET)
> From: Thomas Voegtle <tv at lio96.de>
> Subject: Re: [linux-lvm] lvcreate does not respect units?
> To: LVM general discussion and development <linux-lvm at redhat.com>
> Message-ID: <alpine.LNX.2.00.0912141527510.25221 at er-systems.de>
> Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed
>
> On Mon, 14 Dec 2009, Alasdair G Kergon wrote:
>
>> On Mon, Dec 14, 2009 at 02:45:21PM +0100, Thomas Voegtle wrote:
>>> vgdisplay  Free  PE / Size       6127 / 25.70 GB
>>
>>> root@(none):~# lvcreate -L 25.50g datavolume -n datavolume
>>> lvcreate  Insufficient free extents (6127) in volume group datavolume:
>>> 6528 required
>>
>>> So it seems lvcreate does not respect the unit in power of 1000?
>>
>> Use either:
>>   lvcreate -l 6127
>> or
>>   lvcreate -l 100%FREE
>
>
> Yes, thanks. I know.
>
>
> But I thought you guys would like to hear from bugs?
>
> Why is it possible to use vgdisplay with units power of 1000, if you
> cant use lvcreate with that kind of units?
>
> The lvcreate man page says:
>
>      -L, --size LogicalVolumeSize[bBsSkKmMgGtTpPeE]
>
>
>
> ------------------------------
>
> Message: 6
> Date: Mon, 14 Dec 2009 14:45:19 +0000
> From: Alasdair G Kergon <agk at redhat.com>
> Subject: Re: [linux-lvm] lvcreate does not respect units?
> To: LVM general discussion and development <linux-lvm at redhat.com>
> Message-ID: <20091214144519.GC14009 at agk-dp.fab.redhat.com>
> Content-Type: text/plain; charset=us-ascii
>
> On Mon, Dec 14, 2009 at 03:28:24PM +0100, Thomas Voegtle wrote:
>> Why is it possible to use vgdisplay with units power of 1000, if you
>> cant use lvcreate with that kind of units?
>> The lvcreate man page says:
>>      -L, --size LogicalVolumeSize[bBsSkKmMgGtTpPeE]
>
> For input, it ignores case.
>
> I've not felt particularly happy about changing that as it could change the
> behaviour of existing scripts, and typically, people pay little attention
> to the case of units when they are typing.
>
> But should we change this in time for RHEL6?
>
> Any more opinions?
>
> Alasdair
>
>
>
> ------------------------------
>
> Message: 7
> Date: Mon, 14 Dec 2009 17:21:31 +0100 (CET)
> From: Thomas Voegtle <tv at lio96.de>
> Subject: Re: [linux-lvm] lvcreate does not respect units?
> To: LVM general discussion and development <linux-lvm at redhat.com>
> Message-ID: <alpine.LNX.2.00.0912141719030.31679 at er-systems.de>
> Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed
>
> On Mon, 14 Dec 2009, Alasdair G Kergon wrote:
>
>> On Mon, Dec 14, 2009 at 03:28:24PM +0100, Thomas Voegtle wrote:
>>> Why is it possible to use vgdisplay with units power of 1000, if you
>>> cant use lvcreate with that kind of units?
>>> The lvcreate man page says:
>>>      -L, --size LogicalVolumeSize[bBsSkKmMgGtTpPeE]
>>
>> For input, it ignores case.
>>
>> I've not felt particularly happy about changing that as it could change the
>> behaviour of existing scripts, and typically, people pay little attention
>> to the case of units when they are typing.
>
> Ah, ok, I agree.
>
>
>> But should we change this in time for RHEL6?
>>
>> Any more opinions?
>
>
> Why don't you just do something like si_unit_consistency again, but this
> time for input? Or something similiar?
>
>
>
> ------------------------------
>
> Message: 8
> Date: Thu, 17 Dec 2009 18:41:23 -0500
> From: "Brian J. Murrell" <brian at interlinx.bc.ca>
> Subject: [linux-lvm] brainfart: lilo'd a PV
> To: linux-lvm at redhat.com
> Message-ID: <1261093282.21162.564.camel at pc.interlinx.bc.ca>
> Content-Type: text/plain; charset="utf-8"
>
> I managed to write a lilo block (lilo) to the start of a PV which I have
> on /dev/sda (note, /dev/sda is a whole disk PV, unpartitioned).
>
> However, looking at a number of different PVs here, it appears that the
> LVM data starts at 0x200 with: "LABELONE".  i.e.:
>
> # hexdump -C /dev/sdc3
> 00000000  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
> *
> 00000200  4c 41 42 45 4c 4f 4e 45  01 00 00 00 00 00 00 00  |LABELONE........|
> ...
> # hexdump -C /dev/sdb
> 00000000  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
> *
> 000001b0  00 00 00 00 00 00 00 00  dc 59 0f 10 cf c9 00 00  |.........Y......|
> 000001c0  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
> *
> 00000200  4c 41 42 45 4c 4f 4e 45  01 00 00 00 00 00 00 00  |LABELONE........|
> ...
> # hexdump -C /dev/sda
> 00000000  fa eb 31 12 00 00 4c 49  4c 4f 16 08 10 00 01 00  |..1...LILO......|
> ...
> 00000180  30 e4 cd 13 eb 91 e8 93  fe 44 69 73 6b 20 72 65  |0........Disk re|
> 00000190  61 64 20 65 72 72 6f 72  0d 0a 00 00 00 00 00 00  |ad error........|
> 000001a0  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
> 000001b0  00 00 00 00 00 00 00 00  c6 08 1a 45 cf c9 00 00  |...........E....|
> 000001c0  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
> *
> 000001f0  00 00 00 00 00 00 00 00  00 00 00 00 00 00 55 aa  |..............U.|
> 00000200  4c 41 42 45 4c 4f 4e 45  01 00 00 00 00 00 00 00  |LABELONE........|
>
> Is that correct?
>
> Is there anything I should do to repair this or is my brainfart benign
> to this PV?
>
> Thanx!
>
> b.
>
> -------------- next part --------------
> A non-text attachment was scrubbed...
> Name: not available
> Type: application/pgp-signature
> Size: 197 bytes
> Desc: This is a digitally signed message part
> Url : https://www.redhat.com/archives/linux-lvm/attachments/20091217/49feb21d/attachment.bin
>
> ------------------------------
>
> _______________________________________________
> linux-lvm mailing list
> linux-lvm at redhat.com
> https://www.redhat.com/mailman/listinfo/linux-lvm
>
> End of linux-lvm Digest, Vol 70, Issue 6
> ****************************************
>




More information about the linux-lvm mailing list