Hi,
I have x86 Solaris 10, running on VMWare. I am facing a performance issue on this. There are 5 NGZ running on this server. Whenever netbackup is running on this server, one of the zone started seeing CPU utilization goes as high as 80% and load average to 40-50. Here is example -
-bash-3.2# zoneadm list -icv
ID NAME STATUS PATH BRAND IP
0 global running / native shared
1 zone1 running /zones/zone1 native shared
2 zone2 running /zones/zone2 native shared
3 zone3 running /zones/zone3 native shared
4 zone4 running /zones/zone4 native shared
5 zone5 running /zones/zone5 native shared
-bash-3.2#
It is always zone2, which sees high utilization when netbackup runs. If I run prstatat it tells me it is zpool-zone2/162 which consumes all CPU. "zpool iostat" tells me that IO are very on this pool.
I am still trying to figure it out, but I need another tool to test READs on this filesystem. Instead of waiting for netbackup to run an then I monitor, is there any other command, which can generate lot of READs on /zones/zone2 ?
I could find some dd commands (first write and then read) for Linux, but can't see a valid one for Solaris.
Please suggest.
Thanks