Kibana vs Grafana

Post by **WillemDH** » Wed May 13, 2015 8:18 am

Hello,

Could I get Nagios opinion on this topic. Our Netapp support installed a Grafana server today as it seems Netapp is some kind of a Grafana partner lately. Now I did some research and it seems Grafana is a fork of Kibana. We are in need of a tool that can graph data with a very small interval (1sec)

Should I start learning how to work with Grafana or is the NLS capable of graphing counters etc just like Grafana? If so, would we be able to import the Netapp dashboards from Grafana to NLS?

Grtz

Willem

jdalrymple · Post by **jdalrymple** » Wed May 13, 2015 11:07 am

WillemDH wrote:Hello,

Could I get Nagios opinion on this topic.

I don't want to say I have the "Nagios" opinion - but I may have a useful one nonetheless

WillemDH wrote:Should I start learning how to work with Grafana or is the NLS capable of graphing counters etc just like Grafana? If so, would we be able to import the Netapp dashboards from Grafana to NLS?

I'm going to steal a quote from here because his wording is more precise on the topic than I could blabber out:

@logongas Well in a way it is a fork of Kibana, but it is more like based on the dashboard & panel architecture. None of the panels from Kibana and the elasticsearch query & analytics code is in Grafana. The main focus for Grafana is timeseries and graph panels and graphite is the main focus datasource. Grafana will not be a log analytics dashboard like Kibana. I think having some focus on graphs and timeseries will create a better product than a dashboard that tries to be the dashboard for everything.

But ES will definitely be used as a source for annotations, and to support quick search for metrics keys, as a store for dashboards, panels and settings. MAYBE down the line there might be limited support for some kind of data panel or facet panel based on ES query.

My wording would be something like they're 2 tools that are similar in functionality but their functionality is geared towards totally different consumers and for totally different providers. Kibana is for regurgitating data from ES while Grafana is geared toward regurgitating data from graphite. I don't envision integration one way or the other - and certainly not numeric metrics from NetApp storage into our Nagios "Log" Server product.

My views may not reflect the views of all Nagios employees though - so take it with a grain of salt

Post by **WillemDH** » Mon May 18, 2015 6:43 am

Hmm I was just looking for an autorefresh option for a panel, but I didn't found it. Is it possible to make a panel histogram autorefresh every x seconds?

jolson · Post by **jolson** » Mon May 18, 2015 9:46 am

Willem,

I don't think that auto-refreshing a particular histogram is possible, but have you explored the global auto-refresh option?

2015-05-18 09_45_08-Dashboard • Nagios Log Server.png

Post by **WillemDH** » Mon May 18, 2015 10:05 am

Jesse,

Superb. A per dashboard autorefresh setting will do the trick! Nice.

So I made a do while with my Powershell SendJsonOverTcp function sending some disk related perfmon info to NLS, resulting in a nice near realtime dashboard. If only I could plot multiple field values in one histogram, see my other thread.

EDIT: Just read your response in the other thread. You can close this one too. Looking forward to Kibana 4

Thanks and grtz

Willem

jolson · Post by **jolson** » Mon May 18, 2015 10:18 am

Locking it up.

Nagios Support Forum

Kibana vs Grafana

Kibana vs Grafana

Re: Kibana vs Grafana

Re: Kibana vs Grafana

Re: Kibana vs Grafana

Re: Kibana vs Grafana

Re: Kibana vs Grafana