Performance monitoring with ARGO

monitoring
(Bruce Becker) #1

Currently, ARGO is using Nagios to check the availability of services, and produces great availability reports. From a service management point of view, I interpret this as fulfilling the Service Availability and Continuity Management role - if the service goes down, someone gets informed, and you can write a report saying how much the service has been down, compared to the agreed level.

I would like to be able to do something similar, but for Capacity. From what I’ve seen so far, it seems that ARGO is not providing any metrics related to utilisation - ie, I want to be able to answer the question

  • how much of the resource is available to consume? (if it’s a fixed resource, like storage space or memory)
  • how long are service requests taking?

I read the ARGO probe guide which says that there is room for adding performance metrics to the probes - something explained in the nagios documentation, but I wonder how and where that would be displayed?

Would it be feasible to have it in ARGO?

Would it be feasible to consume nagios data from ARGO and feed it into a dashboard with Grafana or something?

(Yin Chen) #2

A ENVRIplus use case makes use of ARGO for Near Real Time data processing. Check the recent demo video at https://youtu.be/p3UQZkRRWlw

(Kostas Koumantaros) #3

Hi Bruce,

as always there is no easy answer on this and perhaps we should arrange a call to talk about this, when the Argo team submitted the bid we included an option for performance data to be included / consumed by ARGO but this offer was rejected due to budget constrains and as we stand now the team has a full schedule ahead with quite a few tasks/target/commitments adding yet another channel for requests does not help.

Cheers

Kostas