process historical data

[Maxwellb99]

processes.PNG

This is incredible. We are in love with this feature.
Right now we're running this as an active NCPA check every 10 minutes with no thresholds. We'd like to use this for root cause analysis & to contribute the data to our continuous improvement IT process committee.

My question is: What is the best way to access historical status information? Is there a way to access this through the GUI?

to motivate my question. Imagine, a week after an incident, team w wants to know which processes were running between x to y on day z. We'd love to send them the screen capture given above for each discrete interval that we have data for.

top.png

I see in the graph I can get the performance data metrics (process_count, cpu, memory, memory_vms, & memory_rss). As stated above we're looking for a little more verbosity to the data. Also, there's nothing in history, core, etc as it never changes state.

Thanks,
Maxwell Ramirez

[scottwilkerson]

You would want to change the time period on the graph.

Additionally, you can grab the data from the the Nagios API
Help ->REST API Docs -> Objects Reference -> objects/rrdexport

[Maxwellb99]

Hi Scott,

Thanks for your response. This is an area that I'm weak in (rrd, visuals, etc) so you might need to break this down for me barney-style. From what I can see the graph gives me the "performance data" (process_count, cpu, memory, memory_vms, & memory_rss). Similarly for the API call. What we're interested in is exporting the "status information".

If I send them a graph saying 'your CPU was at 100%' they're going to tell me 'I know, that's why the server shut down'. I need to be able to send them the "status information" saying 'PID: Name: Username: Exe: Memory: CPU:' at this time was killing CPU.

Is status information saved off or only Performance Data saved? Can I access this information historically? Am I missing something obvious?

Thanks,
Maxwell Ramirez

[scottwilkerson]

Only performance data is saved per check, however you can reference the state history report for the host/service and see what it said when the state changed (from OK-> CRITICAL or CRITICAL -> OK)