syslog error

[Lorenzo Facello]

Hi all,
in my syslog I find:

Jan 21 13:26:39 rmprsl20 vmunix: vxfs: mesg 056: vx_dataioerr - /dev/vg01/lvol3 file system file data read error
Jan 21 13:29:36 rmprsl20 EMS [1307]: ------ EMS Event Notification ------ Value: "CRITICAL (5)" for Resource: "/storage/events/disks/default/0_3_0_0.0.0" (Threshold: >= " 3") Execute the following command to obtain event details: /opt/resmon/bin/resdata -R 85655696 -r /storage/events/disks/default/0_3_0_0.0.0 -n 85655553 -a

Ok...there are some problem on lvol3...but
What does it mean the EMS's message??
Can you interpretate better this message?
Thanx

[Hoefnix]

The Online diagnostics found a Hardware failure. I did not find your EMS Event yet, but have a look at the url: http://docs.hp.com/hpux/onlinedocs/diag/ems/eme_summ.htm

It looks like a critical error with a disk.
Run the sugetsted command for more info: /opt/resmon/bin/resdata -R 85655696 -r /storage/events/disks/default/0_3_0_0.0.0 -n 85655553 -a

Regards,

Peter

[Fabio Ettore]

Ciao Lorenzo,

probably two messages indicate the same problem, problem on a bad disk.
Peter's suggestion about docs.hp.com is the best suggestion.

Just a question: do bdf and lvdisplay show the same size for /dev/vg01/lvol3?
I have seen that message about vx_dataioerr message when sizes are different between those commands.

As I always say: just a thought....

Best regards,
Ettore

[Luk Vandenbussche]

Lorenzo,

Is it a SCSI or Fiber disk?
We had a while ago the same problem with an MA8000

[Mitra Bhanu Rath]

Hi,
Just run the command as given by EMS to get the detailed error "/opt/resmon/bin/resdata -R 85655696 -r /storage/events/disks/default/0_3_0_0.0.0 -n 85655553 -a"
It is giving some error of "reset" or can say connection reset may.

bhanu

[Olivier Decorse]

Hi Lorenzo.
The only way to make a complete check of your (hard)disks is to read every block of you physical volume, with the "dd" command.
If lvol3 is located on /dev/dsk/c1t15d0, for example (check with lvdisplay -v /dev/vg00/lvol3), just run :
dd if=/dev/rdsk/c1t15d0 of=/dev/null bs=1024k

It just reads all blocks (if=input file)of /dev/rdsk/c1t15d0 (don't forget the "r"dsk), searching for a bad one, writing what is read nowhere (/dev/null).

If the command returns an error, your disk as a problem. If everything is ok, number of read records must be equal to number or write records (36000 in my case) :
36000+0 records in
36000+0 records out

It takes 20 minutes to check a entire 36Go SCSI ULTRA-3 disk.

Olivier.