based on some existing templates and scripts for monitoring Hitachi SAN devices I managed to get
it working for my needs and maybe some other people can benefit as well.
This is working for a Hitachi HUS130 using their SNM2CLI Linux binaries which collects
perfomance data and a perl script making it "cacti-readable".
First, you need to get SNM2CLI working, specifically the "auperform" binary. Setting
the binaries up is a bit tricky, easiest way for me was to use the suggested path oth /usr/stonavm
and export the variables accordingly (it´s documented). Once you can run the binaries, add your
SAN like this: http://www.storit-blog.com/2010/07/31/h ... -part-one/.
Get a read-only user and try to run auperform, the program, that does export perfomance data
about your SAN, ie
Code: Select all
/usr/stonavm/auperform -unit HUSName -auto 5 -count 288 -pfmstatis -path /tmp/hitachi_stats/If you get perfomance data in form of text-files every 5 minutes then that part is working.
For our setup we use a different machine that collects the data so we used SSHFS to connect /tmp/hitachi_stats with our
cacti server. The shell-script that is running does run auperform with "expect" because you have to enter username and password.
Next is to get cacti to grab the output from auperfom and slice it into readable chunks. That´s been done via
the 2 perl scripts you put into the scripts directory. Import Cacti templates and you should be ready to query and
poll your SAN for IO, Cache, CPU, Transfer Rates and Cache Misses. Maybe more once i got the time. Basically all the
information that is in the output of auperfom.
Pictures say more than lots of words, have fun!