topic Re: Urgent: Failure possibly caused by PVG-Strict or Distributed allocation policies when I add mirror in Operating System - HP-UX

Urgent: Failure possibly caused by PVG-Strict or Distributed allocation policies when I add mirror

study unix — Tue, 04 Jan 2011 08:25:43 GMT

Hello Master,
when I add one mirror get the error message as below,"Failure possibly caused by PVG-Strict or Distributed allocation policies",
how to deal with it, I need to create distributed LV, thanks in advance!
# lvcreate -n lv01 -D y -s g /dev/vg16
Logical volume "/dev/vg16/lv01" has been successfully created with
character device "/dev/vg16/rlv01".
Volume Group configuration for /dev/vg16 has been saved in /etc/lvmconf/vg16.conf

# lvextend -l 25 /dev/vg16/lv01 /dev/dsk/c145t0d1 /dev/dsk/c145t0d2
Logical volume "/dev/vg16/lv01" has been successfully extended.
Volume Group configuration for /dev/vg16 has been saved in /etc/lvmconf/vg16.conf
You have mail in /var/mail/root

# lvextend -m 1 /dev/vg16/lv01 /dev/dsk/c143t0d1 /dev/dsk/c143t0d2

Device file path "/dev/dsk/c143t0d1" is an alternate path
to the Physical Volume. Using Primary Link "/dev/dsk/c145t0d1".
Device file path "/dev/dsk/c143t0d2" is an alternate path
to the Physical Volume. Using Primary Link "/dev/dsk/c145t0d2".
lvextend: Not enough free physical extents available.
Logical volume "/dev/vg16/lv01" could not be extended.
Failure possibly caused by PVG-Strict or Distributed allocation policies.

# vgdisplay -v /dev/vg16 |more

--- Volume groups ---
VG Name /dev/vg16
VG Write Access read/write
VG Status available
Max LV 255
Cur LV 1
Open LV 1
Max PV 16
Cur PV 2
Act PV 2
Max PE per PV 2559
VGDA 4
PE Size (Mbytes) 4
Total PE 3070
Alloc PE 25
Free PE 3045
Total PVG 2
Total Spare PVs 0
Total Spare PVs in use 0

--- Logical volumes ---
LV Name /dev/vg16/lv01
LV Status available/syncd
LV Size (Mbytes) 100
Current LE 25
Allocated PE 25
Used PV 2

--- Physical volumes ---
PV Name /dev/dsk/c145t0d1
PV Name /dev/dsk/c143t0d1 Alternate Link
PV Status available
Total PE 2559
Free PE 2546
Autoswitch On
Proactive Polling On

PV Name /dev/dsk/c145t0d2
PV Name /dev/dsk/c143t0d2 Alternate Link
PV Status available
Total PE 511
Free PE 499
Autoswitch On
Proactive Polling On

--- Physical volume groups ---
PVG Name PVG1
PV Name /dev/dsk/c145t0d1
PV Name /dev/dsk/c145t0d2

PVG Name PVG2
PV Name /dev/dsk/c143t0d1
PV Name /dev/dsk/c143t0d2

Re: Urgent: Failure possibly caused by PVG-Strict or Distributed allocation policies when I add mirror

study unix — Tue, 04 Jan 2011 08:53:03 GMT

who can give some support?

Re: Urgent: Failure possibly caused by PVG-Strict or Distributed allocation policies when I add mirror

Torsten. — Tue, 04 Jan 2011 09:00:27 GMT

The important message is

vextend: Not enough free physical extents available.

Re: Urgent: Failure possibly caused by PVG-Strict or Distributed allocation policies when I add mirror

Hakki Aydin Ucar — Tue, 04 Jan 2011 09:02:44 GMT

in the first place , I think you do not have enough space in extension of mirror disks in second lvextend command.

and after all they are not match in size to each other; you used
# lvextend -l 25
# lvextend -m 1

do you have special purpose ?

Re: Urgent: Failure possibly caused by PVG-Strict or Distributed allocation policies when I add mirror

Torsten. — Tue, 04 Jan 2011 09:04:24 GMT

Please post the first ~15 lines of

# lvdisplay -v /dev/vg16/lv01

and you will see you created the LVOL on both PVs and now you try to mirror it to the *same* PVs!

Re: Urgent: Failure possibly caused by PVG-Strict or Distributed allocation policies when I add mirror

study unix — Tue, 04 Jan 2011 09:28:32 GMT

Hello Torsten,
Output as below,
and I will mirror c145t0d1,c145t0d2 to c143tod1,c143t0d2, but c145t0d1 is not the same disk to c143t0d2, why happen the error message ,and how to create one distributed LV ?
# lvdisplay -v /dev/vg16/lv01 |more
--- Logical volumes ---
LV Name /dev/vg16/lv01
VG Name /dev/vg16
LV Permission read/write
LV Status available/syncd
Mirror copies 0
Consistency Recovery MWC
Schedule parallel
LV Size (Mbytes) 100
Current LE 25
Allocated PE 25
Stripes 0
Stripe Size (Kbytes) 0
Bad block on
Allocation PVG-strict/distributed
IO Timeout (Seconds) default

--- Distribution of logical volume ---
PV Name LE on PV PE on PV
/dev/dsk/c145t0d1 13 13
/dev/dsk/c145t0d2 12 12

--- Logical extents ---
LE PV1 PE1 Status 1
00000 /dev/dsk/c145t0d1 00013 current
00001 /dev/dsk/c145t0d2 00012 current
00002 /dev/dsk/c145t0d1 00014 current
00003 /dev/dsk/c145t0d2 00013 current
00004 /dev/dsk/c145t0d1 00015 current
00005 /dev/dsk/c145t0d2 00014 current
00006 /dev/dsk/c145t0d1 00016 current
00007 /dev/dsk/c145t0d2 00015 current
00008 /dev/dsk/c145t0d1 00017 current
00009 /dev/dsk/c145t0d2 00016 current
00010 /dev/dsk/c145t0d1 00018 current
00011 /dev/dsk/c145t0d2 00017 current
00012 /dev/dsk/c145t0d1 00019 current
00013 /dev/dsk/c145t0d2 00018 current
00014 /dev/dsk/c145t0d1 00020 current
00015 /dev/dsk/c145t0d2 00019 current
00016 /dev/dsk/c145t0d1 00021 current
00017 /dev/dsk/c145t0d2 00020 current
00018 /dev/dsk/c145t0d1 00022 current
00019 /dev/dsk/c145t0d2 00021 current
00020 /dev/dsk/c145t0d1 00023 current
00021 /dev/dsk/c145t0d2 00022 current
00022 /dev/dsk/c145t0d1 00024 current
00023 /dev/dsk/c145t0d2 00023 current

Re: Urgent: Failure possibly caused by PVG-Strict or Distributed allocation policies when I add mirror

Torsten. — Tue, 04 Jan 2011 09:31:31 GMT

Look at this:

--- Physical volumes ---
PV Name /dev/dsk/c145t0d1
PV Name /dev/dsk/c143t0d1 Alternate Link
...

PV Name /dev/dsk/c145t0d2
PV Name /dev/dsk/c143t0d2 Alternate Link

So you have 2 physical volumes with alternate pathes (c145t0d1=c143t0d1).

The LVOL is already on these 2 PVs.

Now you need 2 more PVs to mirror the existing 2 PVs.

Re: Urgent: Failure possibly caused by PVG-Strict or Distributed allocation policies when I add mirror

Torsten. — Tue, 04 Jan 2011 09:39:37 GMT

BTW, this is obviously an array (what model?), so it is most likely already a RAID (1 or 5 or whatever), so distribution or mirroring on the same array doesn't make sense at all.

Re: Urgent: Failure possibly caused by PVG-Strict or Distributed allocation policies when I add mirror

Rita C Workman — Tue, 04 Jan 2011 15:27:20 GMT

I fully agree with Torsten, you are looking at the same disk. Don't confuse multiple paths to the same disk.

However, in spite of the considerable respect I have for Torsten, I must disagree with his view on mirroring.
Mirroring is usually a mgmt decision. The data may be best served by mirroring. I do agree that the array itself becomes a single point of failure, but most disk issues are failure of the disk or the power or network pieces to the array.
The one thing I notice when folks mirror, is that they do not watch where they are mirroring to! I have seen where folks have a disk, which is really just a logical disk, and they mirror to another disk that is actually a logical disk sitting on the exact same physical disk going down the same controller.
So...if you are going to mirror - know your array. First check that the disk needs to be mirrored. If you have RAID0+1 (2way mirror), then you don't need to mirror the disk. It was done at the array level.
But if you do need to mirror disk - then pick your mirror disk properly!

Just my 2 cents,
Rita

Re: Urgent: Failure possibly caused by PVG-Strict or Distributed allocation policies when I add mirror

Torsten. — Tue, 04 Jan 2011 20:23:09 GMT

>> So...if you are going to mirror - know your array.

I fully agree.
Thats why I asked for the array model...

Re: Urgent: Failure possibly caused by PVG-Strict or Distributed allocation policies when I add mirror

Rita C Workman — Tue, 04 Jan 2011 20:32:55 GMT

and that's why you sir have the lettuce headband with a star!!

:)

Re: Urgent: Failure possibly caused by PVG-Strict or Distributed allocation policies when I add mirror

Torsten. — Tue, 04 Jan 2011 20:55:22 GMT

;-)

Just to add - I know most of the HP arrays (beginning with AUTORAID, VA, EVA, MSA..., P2000, XP48 to 24k, up to 3PAR and P9500 in future); IMHO it makes no sense for (almost) all of them to mirror a LUN from OS to the same array...

Re: Urgent: Failure possibly caused by PVG-Strict or Distributed allocation policies when I add mirror

study unix — Wed, 05 Jan 2011 01:26:59 GMT

Hi All,
the disk array is HSV210, and I will try to solve it, thanks!
# ioscan -fnC disk
Class I H/W Path Driver S/W State H/W Type Description
===========================================================================
disk 0 0/0/2/0.0.0.0 sdisk CLAIMED DEVICE TEAC DV-28E-N
/dev/dsk/c0t0d0 /dev/rdsk/c0t0d0
disk 1 0/1/1/1.2.0 sdisk CLAIMED DEVICE COMPAQ BD07289BB8
/dev/dsk/c3t2d0 /dev/rdsk/c3t2d0
disk 116 0/4/1/0.4.12.0.0.0.1 sdisk CLAIMED DEVICE HP HSV210