lvol3 of vg00 with mirror stale extents

[likid0]

Hy,

It has been happening lately(3 months), serveral machines have their lvol3 loosing mirror sync and a disk gets all it's PE stale, for example:

[root@sd2:~]# lvdisplay -v /dev/vg00/lvol3
--- Logical volumes ---
LV Name /dev/vg00/lvol3
VG Name /dev/vg00
LV Permission read/write
LV Status available/stale
Mirror copies 1
Consistency Recovery MWC
Schedule parallel
.....
Bad block off
Allocation strict/contiguous
IO Timeout (Seconds) default

--- Distribution of logical volume ---
PV Name LE on PV PE on PV
/dev/dsk/c0t8d0s2 456 456
/dev/dsk/c0t10d0s2 456 456

--- Logical extents ---
LE PV1 PE1 Status 1 PV2 PE2 Status 2
00000 /dev/dsk/c0t8d0s2 00320 current /dev/dsk/c0t10d0s2 00320 stale
00001 /dev/dsk/c0t8d0s2 00321 current /dev/dsk/c0t10d0s2 00321 stale
00002 /dev/dsk/c0t8d0s2 00322 current
.....
00453 /dev/dsk/c0t8d0s2 00773 current /dev/dsk/c0t10d0s2 00773 stale
00454 /dev/dsk/c0t8d0s2 00774 current /dev/dsk/c0t10d0s2 00774 stale
00455 /dev/dsk/c0t8d0s2 00775 current /dev/dsk/c0t10d0s2 00775 stale

It's no big problem with lvsync it's working again. But I don't get why it's happening


some info on the systems:

11.23 ia december 2007 release.
It happens on different servers, follows no pattern, I get no error on syslog/dmesg/disks.
I have been through the disks with dd several times with no errors..

Any idea why this can be happening ?

[Torsten. · Accepted solution]

IMHO it must be a hardware (or config) problem anyhow. Please give more details (e.g. server model, ioscan -fn output, ...).

[likid0]

Hy torsten, thnx for the help!.

I have attached some info on a .txt

I would discard HDW problem because:

It happens on different servers: rx7620, sd64b,
And with different disks: SAN, different scsi HP disk models.

I was looking at a patch problem, but i can't find anything similar..

[Torsten.]

I have an idea ...

Are the boot disks in a ds2120 by change?

Please post

# mptconfig /dev/mpt0

[likid0]

Scan For Devices ...

---- ADAPTER INFORMATION -----------------------------------------------------

Device File : /dev/mpt0
Hardware Path : 0/0/9/1/0/4/0

---- BUS PARAMETERS ----------------------------------------------------------

Initiator SCSI ID : 7
SCSI Bus Rate : Ultra2
SCSI Bus Width : Wide

---- CHANNEL CAPABILITIES ----------------------------------------------------

Req/Ack Offset : 127
Bus Mode : LVD
Quick Arbitration Selection : Disabled
DT Clocking : Enabled
Packetized : Enabled

---- TARGET PARAMETERS -------------------------------------------------------
Target Description Firmware In Use In Use
Id Version Rate Width
------------------------------------------------------------------------------
8 ST373207LC HPC2 Ultra2 Wide
10 ST373207LC HPC2 Ultra2 Wide

------------------------------------------------------------------------------

[Torsten.]

I have seen similar problems when using a speed of ultra320 with a ds2120 - switching down to ultra160 solved this.
But you already have ultra2.
Do you know how long the cables are?

[Torsten.]

For reference the DS2120 quick specs

http://h18006.www1.hp.com/products/quickspecs/12025_div/12025_div.html

[likid0]

I don't know the actual length of the cable, but the distance from the io chassis is your normal sd nextdoor rack

[likid0]

But as I was saying I get the same problem on servers in an rx7620, with ultra160 busses:

Class I H/W Path Driver S/W State H/W Type Description
=======================================================================
ext_bus 0 0/0/0/3/0 c8xx CLAIMED INTERFACE SCSI C1010 Ultra160 Wide LVD A6793-60001
target 5 0/0/0/3/0.5 tgt CLAIMED DEVICE
disk 22 0/0/0/3/0.5.0 sdisk CLAIMED DEVICE HP 73.4GMAX3073NC
/dev/dsk/c0t5d0 /dev/dsk/c0t5d0s2 /dev/rdsk/c0t5d0 /dev/rdsk/c0t5d0s2
/dev/dsk/c0t5d0s1 /dev/dsk/c0t5d0s3 /dev/rdsk/c0t5d0s1 /dev/rdsk/c0t5d0s3
target 6 0/0/0/3/0.6 tgt CLAIMED DEVICE
disk 23 0/0/0/3/0.6.0 sdisk CLAIMED DEVICE HP 73.4GMAX3073NC
/dev/dsk/c0t6d0 /dev/dsk/c0t6d0s2 /dev/rdsk/c0t6d0 /dev/rdsk/c0t6d0s2
/dev/dsk/c0t6d0s1 /dev/dsk/c0t6d0s3 /dev/rdsk/c0t6d0s1 /dev/rdsk/c0t6d0s3
target 7 0/0/0/3/0.7 tgt CLAIMED DEVICE
ctl 23 0/0/0/3/0.7.0 sctl CLAIMED DEVICE Initiator
/dev/rscsi/c0t7d0


:?

[Torsten.]

Are there ds2120 JBODs too?