UNIX for Dummies Questions & Answers

can someone tell me where I can go look for information of commands to check up on veritas volume manager and veritas file system if an alert is received about there being a disk problem??

I need information on what commands to use to perform diagnostic checkups.

this is quite important so any straight forward answers is greatly appreciated. By the way, am using a SunOS 5.8 system

If you have some problem in the Veritas disk, the
system will log this problem in the /var/adm/messages.

I haven't the latest Veritas Version (I have VM 3.1 and FS 3.3.3),
in thats versions, the volume manager don't log by default, you must change the script vxvm-startup2 to add the -x Flags, see man vxconfigd (export MANPATH=$MANPATH:/opt/VRTSvxvm/man/)

When the Volume manager log information, the default file
is:

/var/vxvm/vxconfigd.log

To query the status of Veritas Objetcs, you can use:
vxprint -Aq

In other machine (HP-UX), I use a script to check errors,
you must to modify at your convenience and correct
the paths to the commands.

Code:

#!/bin/ksh

function alarma_VPO {
   server=`uname -n`
   dia_hora=`date +"%d/%m/%Y %H:%M"`
   msg=$1
   /opt/OV/bin/OpC/opcmsg o=OS msg_grp=OS s=critical n=$server_name msg_tex
t="ERROR: $msg ( $dia_hora ) in $server call Unix Support"

}

function checkdetach {
  failed_disks=`vxprint -AQdF '%name %nodarec' | awk '$2=="on" {print " " $1}'`
  failed_plexes=`vxprint -AQpe 'pl_kdetach || pl_nodarec' -F ' %name'`
  failed_volumens=`vxprint -AQvF ' %name' -e "((any aslist.pl_kdetach==true) || (any aslist
.pl_nodarec)) && !(any aslist.pl_stale==false)"`

    if [ ! -z "$failed_disks" ] || [ ! -z "$failed_plexes" ] || [ ! -z "$failed_volumens" ]
    then

       if [ -z "$failed_disks" ]
       then
            msg_txt=" failed disks: $failed_disks"
       fi

       if [ -z "$failed_plexes" ]
       then
            msg_txt=" failed plexes: $failed_plexes"
       fi

       if [ -z "$failed_volumens" ]
       then
             msg_txt="failed volumes: $failed_volumens"
       fi

       alarma_VPO "Volume Manager failures $msg_txt"
    fi
}

vxnotify -f -w 1800 | while read Wword remain_txt
do
    case $Wword in
    waiting) checkdetach;;
    esac
done

exit 0

Regards. Hugo

added code tags for readability --oombera

Solaris Enterprise Volume Manager Reference Manual

vxdisk list will list all the disks and quick status
# vxdisk list
DEVICE TYPE DISK GROUP STATUS
c0t4d0s2 sliced - - error
c0t5d0s2 sliced crash-p rootdg online
c1t0d0s2 sliced data00p datadg online
c1t1d0s2 sliced data01p datadg online

Note "error" on c0t4d0s2, that drive is not actually controlled by Veritas but by DiskSuite. You may see the same on a different drive if you have drives not controlled by Veritas.

vxprint -th will list all the disks, groups, plexes...more detailed listing AND is one you are suppose to do to help with recovery. It should be saved, printed, put into a D/R book.

# vxprint -th
Disk group: rootdg

DG NAME NCONFIG NLOG MINORS GROUP-ID
DM NAME DEVICE TYPE PRIVLEN PUBLEN STATE
V NAME USETYPE KSTATE STATE LENGTH READPOL PREFPLEX
PL NAME VOLUME KSTATE STATE LENGTH LAYOUT NCOL/WID MODE
SD NAME PLEX DISK DISKOFFS LENGTH [COL/]OFF DEVICE MODE

dg rootdg default default 0 926539269.1025.isybase

dm crash-p c0t5d0s2 sliced 3590 17678493 -
dm crash-s c4t1d0s2 sliced 3590 17678493 -

v vol01 fsgen ENABLED ACTIVE 21356672 SELECT -
pl vol01-01 vol01 ENABLED ACTIVE 21359268 CONCAT - RW
sd crash-p-01 vol01-01 crash-p 0 17678493 0 c0t5d0 ENA
sd crash-s-01 vol01-01 crash-s 0 3680775 17678493 c4t1d0 ENA

Disk group: datadg

DG NAME NCONFIG NLOG MINORS GROUP-ID
DM NAME DEVICE TYPE PRIVLEN PUBLEN STATE
V NAME USETYPE KSTATE STATE LENGTH READPOL PREFPLEX
PL NAME VOLUME KSTATE STATE LENGTH LAYOUT NCOL/WID MODE
SD NAME PLEX DISK DISKOFFS LENGTH [COL/]OFF DEVICE MODE

dg datadg default default 50000 926546267.1194.isybase

dm data00p c1t0d0s2 sliced 2159 8378640 -
dm data00s c5t0d0s2 sliced 2159 8378640 -
dm data01p c1t1d0s2 sliced 2159 8378640 -
dm data01s c5t1d0s2 sliced 2159 8378640 -
dm data02p c1t2d0s2 sliced 2159 8378640 -

Thanks a lot guys

just a few question for RTM.

the commands that you suggested, these aren't gon do any damage to the system right? This is simply to give information, not make any changes?. right?

and also, what is it you look for from the output of vxprint -ht in terms of errors and what would you personally do from that point on to rectify the problem?

Those commands only view the status - does not change anything. Follow the link I provide and read the man pages for them.

As far as vxprint -th, all volumes should be active and enabled. If not, then you probably have a problem. It matters what the problem is, how many disks are part of the mirror, if you have hot spares,....each problem will have it's own solution using the same vx commands. See Sun Enterprise SA Guide for a list of error messages, how to do recovery....