Monitor disk-drives that are over 10TB


Getting negative values when doing a snmpwalk towards a server.
And graphs is way off. Disk is over 10TB.

snmpwalk -v2c -c test x.x.x.x HOST-RESOURCES-MIB::hrStorageSize
HOST-RESOURCES-MIB::hrStorageSize.2 = INTEGER: -1610646017

Is there a way around this problem. Or is it just the limitation of SNMP?


That probably is an SNMP limitation. If Iā€™m doing the math right, 10TB in bytes would overflow the 32bit gauge. (somewhere around 4TB would be the max possible to store accurately) You might be able to change the mib2.xml datacollection file for hrStroage* to be a gauge64 instead of a gauge, but if you do that, you will probably need to wipe out the RRD files for the node so they can be rebuilt with the different field type. Make thorough backups if you decide to test this.

Thanks! Can try this in our test-env