Solaris

hi all

i have a problem..

i have this error in /var/adm/messages

Code:

Mar 24 14:50:28 hostname scsi: [ID 107833 kern.notice]   Requested Block: 191034016                 Error Block: 191034016
Mar 24 14:50:28 hostname scsi: [ID 107833 kern.notice]   Vendor: EMC                                Serial Number: 63C46000W   
Mar 24 14:50:28 hostname scsi: [ID 107833 kern.notice]   Sense Key: Unit Attention

I want to know the file system that has these problem
how can i do it ?

Regards
Francesco.

have a look at:

Code:

# iostat -e
# iostat -eE
# df -h
# echo | format

the output is :

Code:

ssd17     Soft Errors: 0 Hard Errors: 16 Transport Errors: 972 
Vendor: EMC      Product: SYMMETRIX        Revision: 5773 Serial No:  
Size: 107.49GB <107485593600 bytes>

but i have a cluster veritas with volume manager...

which are the commands ?



thanks

vxprint might be what you are looking for...

but:
there are many informations missing... which operation has failed? check /var/adm/messages (the 3 lines from above are not enough output!). also you use a EMC as backend storage. so the error can not be on a single disk because the EMC will give you a raid volume to your host... so you might have a problem with your hba or your cables...

Try these,
vxprint -htrpq
Look for disabled volumes.. How many disks (sd) you have per plex, check if disk is mirrored or concated

vxdisk list -ao alldgs
Check all disks are online

I'd assume if you're using VxVM, you'll be using vxdmp...
vxdmpadm listenclosure all
vxdmpadm listctlr all

If you're using power path...
powermt display dev=all

Make sure all your paths are good..

If you identify the failing disk,
vxdisk list <c?t?d?>
Should give you the list of paths for that disk, you can then check with inq, (EMC inquiry) and get the LUN number and then check with the "it's not my problem" people, aka storage team


Hope this helps

Cheers

---------- Post updated at 01:20 PM ---------- Previous update was at 01:15 PM ----------

Aaand veritas cluster...
Check cluster status.
hastatus -sum

Other commands
hagrp, hares, all start with ha...

Or if configured use the java interface, might be easier if you're not familiar with the cluster configuration...