Proactive monitoring of atlassian tool

[Hitendra Chauhan]

Is someone manage to create custom developed proactive monitoring of atlassian tools ?

[Sebastian Kleinholz]

We use Nagios. Probably this link helps a little bit.

https://serverfault.com/questions/106853/best-practices-for-monitoring-a-jira-installation

[Craig Castle-Mead]

Hey Hitendra,

Proactive monitoring could mean a number of things, so depends on exactly what you're looking to solve. Our arsenal includes:

• NewRelic
  • Infrastructure (monitors the ram/disc/cpu etc)
  • APM (deeper insights into the JVM itself)
  • Synthetics (basic HTTP checks from a number of locations to check the service is alive
•  AWS
  • Covers some of the same bases as NewRelic, but we also leverage CloudWatch as it gives us added insights into some of the AWS specifics (RDS read/write throughput Vs thresholds, ELB queues etc)
• ELK (https://www.elastic.co/elk-stack)
  • Get the nginx proxy logs + application logs and send them to a central elasticsearch database, use Kibana to visualize things like application warns/errors over time, usage by username 

Finding correlations between the various tools can help - but you can also get lost in too much data. May be simpler to find an API endpoint you can call, time how long it takes to respond and if it takes longer than your threshold, or, doesn't return the expected data then you likely have an issue. This follows the pattern of monitoring the application itself, and not the servers as your users use the application, not the servers (directly). 

CCM

[Prerak Y Diwan]

Hi Craig,

Hope you are doing well.

For ELK (https://www.elastic.co/elk-stack)

• Get the nginx proxy logs + application logs and send them to a central elasticsearch database, use Kibana to visualize things like application warns/errors over time, usage by username 

Can you point me to a tutorial on how to accomplish this using ELK stack? 

I found video "https://www.youtube.com/watch?v=SQ2pyS6aTpM"

mentioning about the similar topic but did not show the audience how to actually set it up.

Thanks

Prerak

[Pedro Felgueiras]

This is a verry intressing topic and you got a verry intressing presentation here https://pt.slideshare.net/aca_it/monitor-your-atlassian-stack-like-the-nsa

Curently we are using the Google Analytics, JavaMelody and Nagios. We had tried the ELK stack but the logstash was dificult to configure with catalina.out

I think atlassian should provide some guidelines on this topic. 

[Sebastian Kleinholz]

This discussion is related to this topic.