[linux-lvm] Found duplicate PV 7TGP3THphAvROCTtC4OI2gZrvrAGrn15: using /dev/sdah not /dev/sdq Found duplicate PV Qbpd3ov1sd3GHCwCQHUzkI00X53lbcDC: using /dev/sdai not /dev/sdr Found duplicate PV v0YURmuRm51dfY8br0Bx7XxOOIJyV6oa: using /dev/sdt not /dev/sdc Found

[linux-lvm] Found duplicate PV 7TGP3THphAvROCTtC4OI2gZrvrAGrn15: using /dev/sdah not /dev/sdq Found duplicate PV Qbpd3ov1sd3GHCwCQHUzkI00X53lbcDC: using /dev/sdai not /dev/sdr Found duplicate PV v0YURmuRm51dfY8br0Bx7XxOOIJyV6oa: using /dev/sdt not /dev/sdc Found

[Pedro Gonzalez Zamora]

[-- Attachment #1: Type: text/plain, Size: 385 bytes --]

Hi all,

I'm new in LVM when I execute the command "vgscan -v" I get a message below
I don't why this message appear


Found duplicate PV 7TGP3THphAvROCTtC4OI2gZrvrAGrn15: using /dev/sdah not
/dev/sdq
Found duplicate PV Qbpd3ov1sd3GHCwCQHUzkI00X53lbcDC: using /dev/sdai not
/dev/sdr

Anyone Can help me to understand what is the problem and How  can I fix it?

Bests Regards

Pedro G.

[-- Attachment #2: Type: text/html, Size: 578 bytes --]

Re: [linux-lvm] Found duplicate PV 7TGP3THphAvROCTtC4OI2gZrvrAGrn15: using /dev/sdah not /dev/sdq Found duplicate PV Qbpd3ov1sd3GHCwCQHUzkI00X53lbcDC: using /dev/sdai not /dev/sdr Found duplicate PV v0YURmuRm51dfY8br0Bx7XxOOIJyV6oa: using /dev/sdt not /dev/sdc Found

[Bryn M. Reeves]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Pedro Gonzalez Zamora wrote:
> Hi all,
> 
> I'm new in LVM when I execute the command "vgscan -v" I get a message below
> I don't why this message appear
> 
> 
> Found duplicate PV 7TGP3THphAvROCTtC4OI2gZrvrAGrn15: using /dev/sdah not
> /dev/sdq
> Found duplicate PV Qbpd3ov1sd3GHCwCQHUzkI00X53lbcDC: using /dev/sdai not
> /dev/sdr
> 
> Anyone Can help me to understand what is the problem and How  can I fix it?
> 
> Bests Regards
> 
> Pedro G.

Do you have a SAN/fiber/iSCSI environment with multiple paths to the
storage? If so, then you can configure multipathing and filtering to
allow LVM to take advantage of the redundant paths to the storage.

If you don't have a multipath environment then you need to figure out
why these two devices seem to have the same content. Could they have
been copied (dd, some clever storage mirroring thingamy? broken (split)
RAID1 mirror?).

Regards,
Bryn.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iD8DBQFIoXR56YSQoMYUY94RAigRAKCIh7YRDyiXjqK0rIxuBRbUbTHTvACbBoAL
WB4BOU4CoOlPxs95s6OvNBg=
=0GU3
-----END PGP SIGNATURE-----

Re: [linux-lvm] Found duplicate PVs

[Bryan Kadzban]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: RIPEMD160

Bryn M. Reeves wrote:
> broken (split) RAID1 mirror?

Alternately, software RAID1 -- either a dm-raid type setup (most onboard
"RAID controllers", and in fact almost every consumer "RAID" device) or
md?

If the LVM PVs sit on the RAID1 array instead of the member drives, then
you'll get duplicate PVs if you're scanning the member drives themselves
instead of the entire array.  You need to scan the RAID1 array (and of
course it needs to be assembled before you do the scan).

You can control which devices get scanned by editing the "filter" option
in /etc/lvm/lvm.conf -- you can make it only include the RAID array
device (/dev/mapper/whatever or /dev/mdX), instead of the member drives
(/dev/sdXY).

(And this is probably just a special case of the "broken (split) RAID1
mirror" possibility that Bryn was talking about.)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFIoYamS5vET1Wea5wRAzqpAJ4lLzECth88MHROD6T2mahDqMn0jgCfTfPR
1O2q8f2jZnocvQ1VkPCsIHg=
=ygT4
-----END PGP SIGNATURE-----

Re: [linux-lvm] Found duplicate PVs

[Bryn M. Reeves]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Bryan Kadzban wrote:
> Bryn M. Reeves wrote:
>>> broken (split) RAID1 mirror?
> 
> Alternately, software RAID1 -- either a dm-raid type setup (most onboard
> "RAID controllers", and in fact almost every consumer "RAID" device) or
> md?

That's exactly what I meant by a split/broken RAID1 mirror - either two
disks being presented as SCSI/ATA devices when they contain an array or
an array "controller" (fakeraid) using the wrong drivers. Unlikely to be
MD since LVM2 already performs MD component detection (enabled by
default iirc) and will use the MD device in preference to components.

> You can control which devices get scanned by editing the "filter" option
> in /etc/lvm/lvm.conf -- you can make it only include the RAID array
> device (/dev/mapper/whatever or /dev/mdX), instead of the member drives
> (/dev/sdXY).

Shouldn't be needed for MD devices - just make sure you have:

devices {
[...]
    md_component_detection = 1
}

Somewhere in your lvm.conf. As you say, this is most common with dmraid
 since the "controller" is just a regular ATA part with some BIOS goop
and a bit of metadata on the disks but you can get the same sort of
behavior with "real" RAID controllers in some cases.

> (And this is probably just a special case of the "broken (split) RAID1
> mirror" possibility that Bryn was talking about.)

Yup :)

Regards,
Bryn.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iD8DBQFIoZK36YSQoMYUY94RAid3AKDFmEtVBuIBCKOhNJbsC70oVcRdzgCdFRQs
BjW8XPFFatq0y8GpER6GI+c=
=vplQ
-----END PGP SIGNATURE-----

Re: [linux-lvm] Found duplicate PVs

[Pedro Gonzalez Zamora]

[-- Attachment #1: Type: text/plain, Size: 2175 bytes --]

The messages "Found duplicate PVs" disappear when I add this line in
lvm.conf:

      filter = [ "a/sd[a-r]$/" , "r/.*/" ]

But I don't know  if it is a good solution?

Best

Pedro


2008/8/12 Bryn M. Reeves <bmr@redhat.com>

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Bryan Kadzban wrote:
> > Bryn M. Reeves wrote:
> >>> broken (split) RAID1 mirror?
> >
> > Alternately, software RAID1 -- either a dm-raid type setup (most onboard
> > "RAID controllers", and in fact almost every consumer "RAID" device) or
> > md?
>
> That's exactly what I meant by a split/broken RAID1 mirror - either two
> disks being presented as SCSI/ATA devices when they contain an array or
> an array "controller" (fakeraid) using the wrong drivers. Unlikely to be
> MD since LVM2 already performs MD component detection (enabled by
> default iirc) and will use the MD device in preference to components.
>
> > You can control which devices get scanned by editing the "filter" option
> > in /etc/lvm/lvm.conf -- you can make it only include the RAID array
> > device (/dev/mapper/whatever or /dev/mdX), instead of the member drives
> > (/dev/sdXY).
>
> Shouldn't be needed for MD devices - just make sure you have:
>
> devices {
> [...]
>    md_component_detection = 1
> }
>
> Somewhere in your lvm.conf. As you say, this is most common with dmraid
>  since the "controller" is just a regular ATA part with some BIOS goop
> and a bit of metadata on the disks but you can get the same sort of
> behavior with "real" RAID controllers in some cases.
>
> > (And this is probably just a special case of the "broken (split) RAID1
> > mirror" possibility that Bryn was talking about.)
>
> Yup :)
>
> Regards,
> Bryn.
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.7 (GNU/Linux)
> Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org
>
> iD8DBQFIoZK36YSQoMYUY94RAid3AKDFmEtVBuIBCKOhNJbsC70oVcRdzgCdFRQs
> BjW8XPFFatq0y8GpER6GI+c=
> =vplQ
> -----END PGP SIGNATURE-----
>
> _______________________________________________
> linux-lvm mailing list
> linux-lvm@redhat.com
> https://www.redhat.com/mailman/listinfo/linux-lvm
> read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
>

[-- Attachment #2: Type: text/html, Size: 3193 bytes --]

Re: [linux-lvm] Found duplicate PVs

[Bryan Kadzban]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: RIPEMD160

Pedro Gonzalez Zamora wrote:
> The messages "Found duplicate PVs" disappear when I add this line in 
> lvm.conf:
> 
> filter = [ "a/sd[a-r]$/" , "r/.*/" ]
> 
> But I don't know if it is a good solution?

Not likely, but it depends on your hardware setup -- it depends on what
exactly is on those disks.

If you're using fakeraid, then I'd reject sd.* and accept .* (which
would include the /dev/mapper/XXXX fakeraid DM device).

If you're using multipath, then I don't know -- follow the multipath docs.

If you're using LVM on md-raid, then it should be working properly by
default without this change (though md_component_detection must be set
to 1, so you may want to double-check that setting if your system is set
up like this).
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFIoavmS5vET1Wea5wRAxgxAJ0UO3Vkwudc77vS+46V5ER2uiXc+ACgvR2z
wx3Q8uYyUf/tEKAfKqeQUuo=
=TrGx
-----END PGP SIGNATURE-----