Hi Javier,
 
First, just a quick note to let you know that upgrading to 0.8.2 has solved all of my APC UPS graphing problems.  Very handy to be able to do some historical tracking on our UPS's.
 
So on to the next item!  I'm working on building a poller item that determines if a UPS battery is good or bad.  In the case of an APC, the OID is .1.3.6.1.4.1.318.1.1.1.2.2.4.0.  A return value of 1 indicates the battery is ok and a return value of 2 indicates the battery needs to be replaced.  I've built a new poller item based on the included "a APC Output Status" and it looks like this:
 
ID: 10001
Description: a APC Battery Needs Replacing
Name: replace_battery (if I'm reading the documentation right the name would only be important if I wanted to create an RRD column to graph this value)
Poller Command: snmp_status
Parameters: .1.3.6.1.4.1.318.1.1.1.2.2.4.0,1=OK Battery|2=Bad Battery
 
I've added the poller item into the pre-existing APC Poller Group, and within the Poller Group I've told the poller item to use the pre-existing Alarm APC backend (much like a APC Output Status does).  However, when I add a UPS host that has a known bad battery, I don't get any alarm activity such as you would with the UPS switching to battery or losing IP connectivity.
 
Any thoughts on what I've missed?
 
Thanks,
Mike Concannon
 

Reply via email to