problem with vxfs fs migration from (Clariion) VxVM 4.0 Solaris9 to VxVM5.0 Solaris10

[don12]

Hi,

 

I've a problem with migration of the a.m. file system. A LUN was created on Clariion (connected to SUN machine with Solaris9 and VxVM4.0) and then vxfs filesystem. Then I've deported the test1dg disk group from the old machine, and tried to import it and mount on the new one with Solaris 10, VxVM 5.0. I've received the folllowing errors (I'm using PowerPath, DMP is also on but not used):

 

# mount -F vxfs -o largefiles,suid /dev/vx/dsk/testdg01dg/in   /test2
Sep 15 17:25:30 host-tmp vxio: WARNING: VxVM vxio V-5-3-0 voldmp_errbuf_sio_start: Failed to flush the error buffer 600319cad80 on device 0x13800000030 to DMP

Sep 15 17:25:30 host-tmp vxio: WARNING: VxVM vxio V-5-0-2 Subdisk test1dg01-01 block 202688: Uncorrectable read error

UX:vxfs mount: ERROR: V-3-24706: /dev/vx/dsk/test1dg/in no such device or filesystem on it missing one or more devices
#

 

After reboot I can see the following:

 

Copyright 1983-2008 Sun Microsystems, Inc.  All rights reserved.
Use is subject to license terms.
Hostname: host-tmp
NOTICE: VxVM vxdmp V-5-0-34 added disk array DISKS, datype = Disk

NOTICE: VxVM vxdmp V-5-0-34 added disk array CK200052700065, datype = EMC_CLARiiON

NOTICE: VxVM vxdmp V-5-3-1700 dmpnode 312/0x0 has migrated from enclosure FAKE_ENCLR_SNO to enclosure DISKS

Configuring devices.
/dev/vx/rdsk/rootdg/CRASH is clean
NOTICE: VxVM vxdmp V-5-0-112 disabled path 310/0x20 belonging to the dmpnode 312/0x28

NOTICE: VxVM vxdmp V-5-0-111 disabled dmpnode 312/0x28

WARNING: VxVM vxio V-5-0-2 Subdisk test1dg01-01 block 202688: Uncorrectable read error
svc:/system/filesystem/local:default: WARNING: /sbin/mountall -l failed: exit status 1
Sep 15 17:38:34 svc.startd[7]: svc:/system/filesystem/local:default: Method "/lib/svc/method/fs-local" failed with exit status 95.

Sep 15 17:38:34 svc.startd[7]: system/filesystem/local:default failed fatally: transitioned to maintenance (see 'svcs -xv' for details)

Sep 15 17:38:34 dcs: <1619> network initialization failed

Sep 15 17:38:34 dcs: <1643> network initialization failed

Sep 15 17:38:35 dcs: <1681> network initialization failed

Sep 15 17:38:35 dcs: <1686> network initialization failed



host-tmp console login: Sep 15 17:38:35 dcs: <1691> network initialization failed

Sep 15 17:38:36 dcs: <1696> network initialization failed

Sep 15 17:38:36 dcs: <1701> network initialization failed

Sep 15 17:38:36 dcs: <1706> network initialization failed

Sep 15 17:38:36 dcs: <1711> network initialization failed

Sep 15 17:38:36 svc.startd[7]: platform/sun4u/dcs:default failed repeatedly: transitioned to maintenance (see 'svcs -xv' for details)

Sep 15 17:38:36 svc.startd[7]: failed to abandon contract 59: Permission denied





#######

 

Is it a known problem (are there any patches I should install?) or DMP and PowerPath cannot co-exist even if only PowerPath is used. Can I simply deactivate DMP and leave PowerPath working? How can I do it?

 

Any suggestions?

Thanks!

[bsobek]

Hi,

could you give more information?

- vxdg list testdg01dg

- vxprint -g testdg01dg

- fstyp /dev/vx/dsk/testdg01dg/in  

are all volumes enabled?

thanks!

greets

Björn 

[don12]

Hi Bjorn!

Here are the logs you've asked for, thanks in advance for any suggestions, ...

Despite of these "Uncorrrectable read errors", this fs can be mounted without any problems on the old machine with Solaris9.

Regards, 

Piotr

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

 root@host-tmp# vxdg list test1dg
Group:     test1dg
dgid:      1221491819.27.be5a
import-id: 1024.8
flags:
version:   110
alignment: 8192 (bytes)
ssb:            on
detach-policy: global
dg-fail-policy: invalid
copies:    nconfig=default nlog=default
config:    seqno=0.1129 permlen=1481 free=1476 templen=3 loglen=224
config disk emcpower0c copy 1 len=1481 state=clean online
log disk emcpower0c copy 1 len=224
root@host-tmp# vxprint -g test1dg
TY NAME         ASSOC        KSTATE   LENGTH   PLOFFS   STATE    TUTIL0  PUTIL0
dg test1dg      test1dg      -        -        -        -        -       -

dm test1dg01    emcpower0c   -        202704   -        FAILING  -       -

v  in           fsgen        ENABLED  202704   -        ACTIVE   -       -
pl in-01        in           ENABLED  202704   -        ACTIVE   -       -
sd test1dg01-01 in-01        ENABLED  202704   0        -        -       -
root@host-tmp# fstyp /dev/vx/dsk/test1dg/in
vxfs
WARNING: VxVM vxio V-5-0-2 Subdisk test1dg01-01 block 201680: Uncorrectable read error
WARNING: VxVM vxio V-5-0-2 Subdisk test1dg01-01 block 201680: Uncorrectable read error
WARNING: VxVM vxio V-5-0-2 Subdisk test1dg01-01 block 201680: Uncorrectable read error
WARNING: VxVM vxio V-5-0-2 Subdisk test1dg01-01 block 202192: Uncorrectable read error
WARNING: VxVM vxio V-5-0-2 Subdisk test1dg01-01 block 202192: Uncorrectable read error
WARNING: VxVM vxio V-5-0-2 Subdisk test1dg01-01 block 202192: Uncorrectable read error
root@host-tmp# ls -al /dev/vx/dsk/test1dg/in
lrwxrwxrwx   1 root     root          44 Sep 22 14:09 /dev/vx/dsk/test1dg/in -> /devices/pseudo/vxio@0:test1dg,in,107000,blk
root@host-tmp#
root@host-tmp# /opt/VRTS/bin/fstyp -v /dev/vx/dsk/test1dg/in
vxfs
magic a501fcf5  version 6  ctime Mon Sep 15 15:36:51 2008
logstart 0  logend 0
bsize  8192 size  12669 dsize  0  ninode 12669  nau 0
defiextsize 0  ilbsize 0  immedlen 96  ndaddr 10
aufirst 0  emap 0  imap 0  iextop 0  istart 0
bstart 0  femap 0  fimap 0  fiextop 0  fistart 0  fbstart 0
nindir 2048  aulen 32768  auimlen 0  auemlen 1
auilen 0  aupad 0  aublocks 32768  maxtier 15
inopb 32  inopau 0  ndiripau 0  iaddrlen 1 bshift 13
inoshift 5  bmask ffffe000  boffmask 1fff  checksum edd67252
oltext1 11  oltext2 1796  oltsize 1  checksum2 0
free 5561  ifree 0
efree  1 0 0 3 2 2 3 1 2 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
root@host-tmp#

[bsobek]

Hi,

thanks. We had the same error message on a solaris host. in our case, we had a duplicated major/minor number.

cd /dev/vx/rdmp

ls *emcpower0*

search for the same major/minor number in the device tree.

do you have a running vcs on this system?

one way to solve the problem is (we had opened a case for this problem and the supporter told us this way):

- freeze all servicegroups

- rm -rf /dev/vx/rdmp/* 
- rm -rf /dev/vx/dmp/*

- mv /etc/vx/disk.info /etc/vx/disk.info.old (the same for the other .info files in this directory)
- devfsadm -C

- vxconfigd -k

- vxdctl enable

- unfreeze all servicegroups

You have to do this on all cluster-nodes.  

an other was would be to reboot the system with reconfiguration. I would prefer this method for none cluster environments because you are killing and restarting the vxconfigd.

Greets

Björn

[don12]

Thanks Bjorn!

This helped a little, but only refreshed the configuration, problem still exists.

One thing I've noticed is that when tpdmode is set to native on the destination host (M4000) I'm able to recognize the device on the array and initialize the disk. If it is set to pseudo - it's not possible: I mean it's neither possible to import the group deported from the source host (SF480R) nor initialize a new disk on the array.

So being aware of this, I'm able to do a data migration, however this required data copy from the disks initialized on the source host to the disks initialized on the destination host. Anyway the root cause is still unknown for me.

Any other ideas?

Best Regards, 

Piotr

[The source and destination hosts are not in a cluster config, these are separate machines]