Prod and DR servers have R1/R2s assigned and VxVM is being used on the server. All SRDF related operations are performed at the host level either from Prod or DR node. Ideally, all SRDF devs should be part of Veritas DG as well as symdg. What happens when you perform the DR test, and after performing split or failover find out that wasn’t the case. Newly added disks weren’t in VxDG but not in symdg. Data corrpution or failed DR.
Below is a small utility script my clever colleague Sean Hall put together to verify that:
a) All disks in Veritas disk group are infact in symmetrix DG and are replicated (in sync)
b) That all disks in symm DG are syncronised
#!/bin/ksh
VeritasDG=sharesp1
symdg=sharesp1
SYMINQ=/usr/symcli/bin/syminq
SYMINQOUT="/tmp/$(basename $0)_syminq.out"
/usr/symcli/bin/syminq >${SYMINQOUT} 2>&1
# First check is that all disks in Veritas are replicated
for volume in `vxdisk -g $VeritasDG list | grep -v DEVICE | awk '{ print $1 }'`
do
disk=`grep -w $volume $SYMINQOUT | awk '{ print $6 }' | cut -c4-7`
status=`symrdf -g $symdg query | grep ^DEV | grep $disk | awk '{ print $12 }'`
print "${volume}:${disk}:${status}"
done
# 2nd check is that all disks in symm DG are syncronised
symrdf -g $symdg query | grep ^DEV
No comments yet.