LVM_IO_FAIL - I/O Error Detected by LVM

[george bagley]

I am getting the following in errpt

LABEL: LVM_IO_FAIL
Resource Name: LVDD
Description: I/O ERROR DETECTED BY LVM
PHYSICAL VOLUME DEVICE MAJOR/MINOR: 002D 2008
ERROR CODE AS DEFINED IN sys/errno: 5
BLOCK NUMBER: 72032256
LOGICAL VOLUME DEVICE MAJOR/MINOR: 002E 0006
SENSE DATA:
0004 4B20 0004 6909 281A 1A09 0000 0000 0000 0000 0004 6909 DE4E 9084
0000 0000 0000 0000

I am getting this error every few seconds. (approx 100 every 3 hours)
I am "guessing" that the LVM is finding bad blocks and is relocating
the data to another block. Is this correct?
I would also like to know how to translate major/minor numbers into
pdisks and logical volumes as applicable.

I would like to know point 2 as I am getting the errors on two drives
at present and I am putting six new drives into the array within the
the next ten days. If I can determine which drives are failing, I
will order an extra two drives and replace the failing drives at the
same time. (I am doing a full rebuild from tape of this volume
group.)

The drives for this volume group are SSA in MOD600 towers.

I am running AIX 4.3.3.0.8

Any info would be appreciated.

George Bagley
Systems Engineer, DBA

[Nicholas Dronen]

george bagley <george_...@hotmail.com> wrote:

> I am getting the following in errpt

> LABEL: LVM_IO_FAIL
> Resource Name: LVDD
> Description: I/O ERROR DETECTED BY LVM
> PHYSICAL VOLUME DEVICE MAJOR/MINOR: 002D 2008
> ERROR CODE AS DEFINED IN sys/errno: 5
> BLOCK NUMBER: 72032256
> LOGICAL VOLUME DEVICE MAJOR/MINOR: 002E 0006

> I am getting this error every few seconds. (approx 100 every 3 hours)

> I am "guessing" that the LVM is finding bad blocks and is relocating
> the data to another block. Is this correct?

Possibly. That's sufficiently frequent to indicate a failing
drive, though.

> I would also like to know how to translate major/minor numbers into
> pdisks and logical volumes as applicable.

# physical volume device major/minor
$ echo 'ibase=16; 2D; 2008' | bc
45
8200

# logical volume device major/minor
$ echo 'ibase=16; 2E; 06' | bc
46
6

I'm skeptical of the minor number for the physical device.
Nevertheless, look in /dev/ for a device with a major number
of 45 and a minor number of 8200 (?) for the physical volume;
the logical volume -- the numbers for which are much more
reasonable to me -- has a major number 46 and minor number 6.

Just so you know how to find these devices: the major number
is the first in the comma-separated pair which occurs where
you usually see the file size in the output of ls -l.

$ ls -l /dev/hdisk0
brw------- 1 root system 12, 1 Dec 19 2000 /dev/hdisk0

The major number of hdisk is is 12; its minor number is 1.

Regards,

Nicholas Dronen

P.S. I'm glad to hear you're working; not so for your drives. :-)

[george bagley]

Thanks for that

I still have a problem though

The major and minor number of the physical device point to a raid
array with 16 odd disks in it.
The logical volume indicated is spread over the disks in the array. I
have run diags on the disks and no errors/problems are being flagged.

Is there a way to interpret the sense data so that I can narrow it
down to a pdisk??

Thanks again

George

[Red Alert]

george_...@hotmail.com (george bagley) wrote in message news:<ad097928.01080...@posting.google.com>...

George,
are these errors preceded in the error log by SSA related messages?
If not, I don't think this is a hardware problem and the error is being
recognised by the LVM before anything is passed to the raid manager.

Is it a raw logical volume that the errors are logged against? If so,
what is accessing it? Is 72032256 a valid block number?

Regards,

[naresh1...@gmail.com]

hai team
i am also getting those lvm error
how can i resolve it


LABEL: LVM_IO_FAIL
IDENTIFIER: E86653C3

Date/Time: Fri Apr 4 22:23:48 IST 2014
Sequence Number: 1938617
Machine Id: 0004AB08D700
Node Id: racserver
Class: H
Type: PERM
WPAR: Global
Resource Name: LVDD
Resource Class: NONE
Resource Type: NONE
Location:

Description
I/O ERROR DETECTED BY LVM

Probable Causes
POWER, DRIVE, ADAPTER, OR CABLE FAILURE

Recommended Actions
RUN DIAGNOSTICS AGAINST THE FAILING DEVICE

Detail Data
PHYSICAL VOLUME DEVICE MAJOR/MINOR
8000 001A 0000 0000
ERROR CODE AS DEFINED IN sys/errno.h
5
BLOCK NUMBER
227147776
LOGICAL VOLUME DEVICE MAJOR/MINOR
8000 000A 0000 000D
PHYSICAL BUFFER TRANSACTION TIME
0
RESIDUAL COUNT
131072
NUMBER OF BLOCKS
131072
I/O TYPE
USER DATA
SENSE DATA
0000 0000 0006 C500 0004 AB08 0000 D700 0000 0136 A72A 567F 0004 AB08 97BF 6386
0000 0000 0000 0000

[Lőrinczy Zsigmond]

On 2014.04.10. 9:58, naresh1...@gmail.com wrote:> Description

> I/O ERROR DETECTED BY LVM
>
> Probable Causes
> POWER, DRIVE, ADAPTER, OR CABLE FAILURE
>
> Recommended Actions
> RUN DIAGNOSTICS AGAINST THE FAILING DEVICE

You seem to have hardware-problem. Ask the nearest IBM representative
for help.