Integrate InfluxDB / Grafana into Nagios XI

[WillemDH]

Wouldn't it be nice if InfluxDB / Grafana was somehow integrated in Nagios XI. I've played around with Grafana a lot lately and it is truly awesome for debugging etc. I made a Reactor action to start gathering perfmon counters for a Nagios host. http://outsideit.net/naf-windows-perfmon-to-influxdb/
I think real-time monitoring is 'the future' and is already available in some competing monitoring products I'm not gonna mention here. It's perfectly possible to setup our own Grafana frontent, but imho it would be nice to see it integrated in XI somehow.

[tmcdonald]

Can always bring it up to the developers I suppose. I'm not really familiar with either of those technology stacks, but they look cool. What benefits specifically would they provide?

[mopp]

Can always bring it up to the developers I suppose. I'm not really familiar with either of those technology stacks, but they look cool. What benefits specifically would they provide?

Did you hear anything from the developers?
Is Grafana on the roadmap?
From my understanding Grafana can read data from Graphite, InfluxDB and OpenTSDB. There is no direct connection to RRD files as far as I know.

One benefit of Grafana would be to read data from multiple Nagios server. So you can have a central dashboard for performance graphs. This is not possible with the graphexplorer.

[WillemDH]

This thread is made to add the Grafan webpage under tools. Grafana is as far as I know not on the nagios roadmap. I set up my own Grafana server.

[tmcdonald]

I was waiting to get a use case before I brought this to the devs since they usually will ask for one. But I think Willem is right, it isn't on the map since we don't use Graphite, InfluxDB or OpenTSDB anywhere in our stack.

[awilson]

There is a plugin, Graphios, to send Nagios performance data to Graphite. Then you can use Grafana for visualization. I haven't tested it yet, but it appears conceptually sound so far.

One use case would be to provide a query of bandwidth utilization from a data center as a dashboard item and then let users drill down to the component sources. Adding NNA to the stack, would enable interesting analyses.

Alan

[tmcdonald]

Thanks for the tip @awilson! @WillemDH if you have a good use case or benefits I can bring this to the devs, but without that I don't think they're going to spend time on this.

[WillemDH]

Trevor,

Well I think the use case is very clear. Our 'end-users' want 1-to-10 second interval graphs. RRD, Highcharts, PNP4nagios I've never seen it do granularity of less then 1 minute. Now I personally know 5+ competing products who have already integrated InfluxDB into their monitoring products.

You know I am very invested into Nagios XI. Believe me you guys want to look at InfluxDB asap. It opens up a world of possibilities and the CPU load caused by InfluxDB and Telegraf is very very small.

I wouldn't insist so much if I didn't think integrating InfluxDB would be one of the best choices you guys would make to ensure Nagios superiority.

It's better then rrd's in many ways.

Grtz

Willem

[tmcdonald]

Got your PM, the devs are discussing this and you should have heard back from Scott and Jake. Mind if we close this up since the conversation is continuing elsewhere?

[WillemDH]

Sure, pls close.