I am in total agreement with this request!
We are monitoring numerous PDU's and to have it built into NPM would be fantastic. Instead, as you have alluded to, we have to build custom properties add custom poller objects and build custom views. This, I thought, would be included within NPM since most all corporations/enterprises have PDU's that need to be monitored.
I am looking at a third party service to complete this task due to it being way to labor intensive to build everything that is needed to be functional.
I second the request for PDU's to be added to NPM!
Any news about the first request?
No update yet!
i asked this two years ago from SW, no traction
last year i tried out a Raritan product called DC track and the sister product PowerIQ. we licked and but it all our data centres
- DC track manages our entire data center , from inventory tracking to dynamic maps, to tracing the path from user to the server. we can even enter in our procurement details and it reminds us when renewal dates come up and the POC names. For new people to the firm, it even allows them to drill down from building, floor, row , rack and to the actual device
- PowerIQ integrates with any 3rd party PDU, UPS, generator etc. it gives us reporting and alerts ,
we also use their asset monitor trips. no equipment is touched without us knowing. it ties into our SW help desk ticket system
Could not find anything more recent for PDU posts so sorry for hijacking your post.
OK so we finally caught up on the technology and now all our PDU's in each Cab are sitting on the network also. I can now receive alerts and alarms etc via SNMP/SMTP. I've added them into my Network Diagram in Solarwinds for monitoring. The bad news is, due to the way the power is configured they are giving me nightmares. My question is, would it be possible to create a dependency between the two pdu's confirming ones power status to the other?
The reason why I need this is, each CAB contains two PDU's. The left hand PDU is on the UPS and the right hand PDU is direct power. With the servers having dual power supplies, they receive power from each of the PDU's, when we have our monthly generator test for the building at 7:30am and switched back at 6:00pm , the right hand PDU in each CAB drops power for a second and I get alarms for each of them and every nodes PSU2 effected, around 500 alerts emails and texts for less than 60 seconds switch over. I was wondering if I could create some kind of conditional rule that would stop this to some degree at least, without risk of missing a genuine fault.
a) If its on X day of the month and the time is x and your partnered PDU has not complained. Don't send an Alert Out, just log it
b) Do nothing if again, the CABs other pdu on UPS is alive and not warning. Poll again in five minutes see if everything is back to normal.
c) Create some kind of dependency where the PDU on UPS is polled for its health by the None UPS PDU and as long as one says all is well and I'm not currently running on batteries, suppress your alert.
Has anyone got any ideas how and if this is achievable? I'm sorry if this is confusing in description. Welcome to the world of NHS
There may be a few options for you:
- there is an option to alerting only during specific times or the option to only send the notifications during specific times
- monitor both PDUs, but only alert on the UPS PDUs in general and potentially only alert when paired device has been down > 10 minutes
- When building the alert, use the Advanced drop down on the trigger condition & reset condition to link the two PDUs together. Only send the alert if both are down, or on batteries
- use groups to show best and only alert when whole group is down