meta data write error

[dictum9]

I am getting the following in syslog, any idea what it means? Seems like a hardware issue.

It also caused LVM problems. (see below)




May 8 15:01:03 sd17a vmunix: msgcnt 1 vxfs: mesg 037: vx_metaioerr - vx_write_default - /dev/vgoracle/lvol1 file system meta data write error in block 530776
May 8 15:01:03 sd17a vmunix: msgcnt 2 vxfs: mesg 031: vx_disable - /dev/vgoracle/lvol1 file system disabled
May 8 15:01:04 sd17a vmunix: msgcnt 3 vxfs: mesg 017: vx_write_default - /home/oracle file system inode 24442 marked bad incore
May 8 15:01:04 sd17a vmunix: ?vxfs msgcnt 3 offset 0x00000000 81a4 1 c8 c9
May 8 15:01:04 sd17a vmunix: ?vxfs msgcnt 3 offset 0x00000010 0 1be0d67 44475918 26f43
May 8 15:01:04 sd17a vmunix: ?vxfs msgcnt 3 offset 0x00000020 445f956f 5849a 445f956f cb089
May 8 15:01:04 sd17a vmunix: ?vxfs msgcnt 3 offset 0x00000030 30008 0 0 0
May 8 15:01:04 sd17a vmunix: ?vxfs msgcnt 3 offset 0x00000040 6f94 3ecdf094 0 a72a
May 8 15:01:04 sd17a vmunix: ?vxfs msgcnt 3 offset 0x00000050 1000000 0 b2d10 8
May 8 15:01:04 sd17a vmunix: ?vxfs msgcnt 3 offset 0x00000060 0 0 0 0
May 8 15:01:04 sd17a vmunix: ?vxfs msgcnt 3 offset 0x00000070 0 0 0 0
May 8 15:01:04 sd17a vmunix: ?vxfs msgcnt 3 offset 0x00000080 0 0 0 0
May 8 15:01:04 sd17a vmunix: ?vxfs msgcnt 3 offset 0x00000090 0 0 0 0
May 8 15:01:04 sd17a vmunix: ?vxfs msgcnt 3 offset 0x000000a0 0 0 0 0
May 8 15:01:04 sd17a vmunix: ?vxfs msgcnt 3 offset 0x000000b0 0 0
May 8 15:01:05 sd17a vmunix: Synchronous Page I/O error occurred while paging to/from disk
May 8 15:01:05 sd17a vmunix: device 0x00000000, 1K block #2626076, 0 bytes of 4096 bytes not paged, page addr =15018031.4000000002487000, bflag =1f, bp = e0000001a5fa8400
May 8 15:01:13 sd17a vmunix: device 0x00000000, 1K block #733816, 0 bytes of 4096 bytes not paged, page addr =16730831.400000000009e000, bflag =1f, bp = e0000001aebc1c40
May 8 15:01:14 sd17a rexecd[11352]: unable to change to users homedir:/home/oracle
May 8 15:01:14 sd17a rexecd[11353]: unable to change to users homedir:/home/oracle
May 8 15:01:13 sd17a vmunix: Synchronous Page I/O error occurred while paging to/from disk
May 8 15:01:19 sd17a rexecd[11355]: unable to change to users homedir:/home/oracle
May 8 15:01:30 sd17a rexecd[11361]: unable to change to users homedir:/home/oracle
May 8 15:01:31 sd17a vmunix: Synchronous Page I/O error occurred while paging to/from disk
May 8 15:01:31 sd17a vmunix: device 0x00000000, 1K block #214776, 0 bytes of 4096 bytes not paged, page addr =3bfc0031.c000000004789000, bflag =1f, bp = e0000001aebc1c40
May 8 15:01:40 sd17a rexecd[11364]: unable to change to users homedir:/home/oracle
May 8 15:02:14 sd17a rexecd[11386]: unable to change to users homedir:/home/oracle





vgchange -a y /dev/vgoracle

vgchange: Warning: Couldn't attach to the volume group physical volume "/dev/dsk/c10t1d1":
Cross-device link
vgchange: Cross-device link
vgchange: Warning: couldn't query physical volume "/dev/dsk/c10t1d1":
The specified path does not correspond to physical volume attached to
this volume group

[Steven E. Protter · Accepted solution]

Shalom etc,

Same as your other thread.

Disk is dead or unreachable.

cstm/mstm/xstm

Diagnose and repleace.

SEP

[dictum9]

I see no problems with either vgdisplay, lvdisplay, pvdisplay or ioscan -fnC disk, or dmesg.
All appears as normal.
Are you saying that the only thing I can do is run xstm?

[Pradeep_3]

check in lvmtab file for the disk "/dev/dsk/c10t1d1"
If volume group is disable then how can you do vgdisplay,pvdisplay or lvdisplay.

[Court Campbell]

Pradeep, in response to what you said, the vgdisplay, lvdisplay, an pvdisplay is in memory. You do not actually query the disk pvra/vgra. it would only get updated in memeory if you did a manual vhcange -a y .

[Bill Hassell]

There are several disk error messages:

/dev/vgoracle/lvol1 file system disabled
unable to change to users homedir:/home/oracle
Couldn't attach to the volume group physical volume /dev/dsk/c10t1d1

Most likely everything associated with /dev/dsk/c10t1d1 is dead or full of errors. You can verify this with:

dd if=/dev/rdsk/c10t1d1 of=/dev/null bs=256k

This will scan the disk for read errors.