# # Just in case of problems, let’s not hang Nagios
# $SIG{’ALRM’} = sub {
# print “ERROR: No snmp response from $hostname (alarm timeout)\n”;
# exit $ERRORS{’UNKNOWN’};

and it seems to work while doing the manual checks. Since it’s a distributed system when DMS sends the info to CMS it still errors out

Other NetApp gotchas are that they change the units they report metrics in between releases. e.g. latency for reads and write is reported in milliseconds before 7.3, microseconds after. And should not even be collected on early 7.3 code, due to bugs.
This is some of the stuff that LogicMonitor automates and saves lots of time on.

To get valid data between 2 and 4TB, you need to use unsigned integers. Larger volume/aggregate sizes are also supported in the v1 and v2c MIB but they are split into a low and high value, so you need to combine them. If you look through the MIB, you will see entries like:

dfHighAvailKBytes
dfLowAvailKBytes

in the dfTable section so you need to shift the first value left 32 bits and add them together.

I was hoping someone had already put this into the plugin but I guess not yet. I only started looking at this a couple of days ago when I got an insane response for a 6TB volume.

The systems that are <10ms away checks are coming back just fine
Anything further away I’m doing -t 150.
During manual check checks come back fine, but automatic checks still come back timed out.
Any ideas?

Also volumes over 10TB come back with funky numbers, free space is OK, but the totals are incorrect.

$used = unpack(”I”, pack(”i”, $used));