Hi,
Got a weird issue with UPS data collection for devices that are in ‘Eco Mode’. (I’m not a fan of the mode, but some customers run them this way.)
In brief, it means the UPS is running on internal bypass with the inverter synchronised but off-load. When there is a failure then it starts being a UPS again, quickly. This means that 99.9% of the time it has near-zero “Input Current” because the load is being fed from the bypass line. This seems to cause Observium to cease polling the values for the Input Currents (they are zero, or near-zero) on some of the three phases.
The most annoying side effect is that they disappear and re-appear (depending if the value is zero or 0.1A) during each detection run, so the custom sensor values seem to get reset. Causing random alarms when the values drop to zero again.
[cid:image001.png@01D26FDB.CB82BEB0]
Interestingly, it does not seem to have this issue with the Bypass Current which (in this device) isn’t metered and therefore is always zero, but always gets found OK.
There is residual current on Input Phase 2 and Input Phase 3 but rarely does it show anything on Input Phase 1 (until the UPS goes active).
I hope that this all makes a small amount of sense - attached is a debug discovery and a debug poll of the sensors module for this example device. (The last run I did it found Phase 2 and Phase 3.)
I have also attached an SNMP walk of the device. I’ve done this several times and I always get all the OIDs responded to, but phases 1 and 2 are often zero:
root@:~# snmpwalk <blah> 1.3.6.1.2.1.33.1.3.3.1.4 iso.3.6.1.2.1.33.1.3.3.1.4.1 = INTEGER: 0 iso.3.6.1.2.1.33.1.3.3.1.4.2 = INTEGER: 1 iso.3.6.1.2.1.33.1.3.3.1.4.3 = INTEGER: 2 root@:~# snmpwalk <blah> 1.3.6.1.2.1.33.1.3.3.1.4 iso.3.6.1.2.1.33.1.3.3.1.4.1 = INTEGER: 0 iso.3.6.1.2.1.33.1.3.3.1.4.2 = INTEGER: 1 iso.3.6.1.2.1.33.1.3.3.1.4.3 = INTEGER: 2 root@:~# snmpwalk <blah> 1.3.6.1.2.1.33.1.3.3.1.4 iso.3.6.1.2.1.33.1.3.3.1.4.1 = INTEGER: 0 iso.3.6.1.2.1.33.1.3.3.1.4.2 = INTEGER: 1 iso.3.6.1.2.1.33.1.3.3.1.4.3 = INTEGER: 2 root@:~# snmpwalk <blah> 1.3.6.1.2.1.33.1.3.3.1.4 iso.3.6.1.2.1.33.1.3.3.1.4.1 = INTEGER: 0 iso.3.6.1.2.1.33.1.3.3.1.4.2 = INTEGER: 0 iso.3.6.1.2.1.33.1.3.3.1.4.3 = INTEGER: 1 root@:~# snmpwalk <blah> 1.3.6.1.2.1.33.1.3.3.1.4 iso.3.6.1.2.1.33.1.3.3.1.4.1 = INTEGER: 0 iso.3.6.1.2.1.33.1.3.3.1.4.2 = INTEGER: 0 iso.3.6.1.2.1.33.1.3.3.1.4.3 = INTEGER: 1
So – bug or UPS card fail? Any ideas? Cheers!
Robert Williams Custodian Data Centre Email: Robert@CustodianDC.com http://www.CustodianDC.com
