Aggregate graphs metrics, in my case for PDU's

Help
1

Can LibreNMS aggregate into one graph power readings from n+1 individual PDUs? It would be useful to see an aggregated graph for a complete data center, as well as, and probably more importantly, aggregate two PDU’s in the same cabinet that are being used in an active/active redundancy senario. IE, both PDU’s are being used, and both have a capacity of 30A (24amps usable) and I want to be sure I am never going past the capacity of a single circuit so that the removal of power from one PDU will allow the complete load to exists on the second. In this scenario I want to know that my total load is at or under 24 amps. I want to alert at ~20-24 amps, and I want to see this visually on one graph per rack. Knowing the power that exist on each circuit is nice, but not useful for alerting, and not convenient for power capacity planning.

Or simply put, to abstract the need to care about my power needs, how do I create one graph that sums the same metric from two to discrete but identical (or similar) devices?

Not sure if this is a request for help for an undocumented feature, or if this is a feature request.

3 Likes
2

This isn’t a feature at present so will need to be added.

3

Thanks for the answer laf. Is this on the road map? Seems like this would be desirable for a fair amount of use cases. I am not familiar with the guts of LibreNMS so I am not sure how hard it is to implement. I am trying to pair down the number of monitoring and alerting applications and consolidating on LibreNMS for SNMP based metric and alerts makes sense, but this is a wrinkle for me. Any insight into the chances of this becoming a feature soon or not appreciated.

4

Features get added when someone needs to fix something or has the time so it’s impossible to say when this would get added.

We already have support for combined port graphs so it might not be that difficult - but without looking and starting the code it’s hard to say.