Galaxy Monitoring with Telegraf and Grafana

Contributors

Authors: Nate Coraor Björn Grüning Simon Gladman Helena Rasche

Questions

• How to monitor Galaxy with Telegraf

• How do I set up InfluxDB

• How can I make graphs in Grafana?

• How can I best alert on important metrics?

Objectives

• Setup InfluxDB

• Setup Telegraf

• Setup Grafana

• Create several charts

Telegraf, InfluxDB, and Grafana

General purpose tools for monitoring systems and services.

Dataflow:

• Galaxy produces data
• Telegraf consumes and buffers it, before sending it to
• InfluxDB which stores the data
• And Grafana is used to visualise it

Speaker Notes

• Monitoring in Galaxy is easy to setup.
• Galaxy produces data, which is consumed by telegraf.
• telegrafends data to Influx DB.
• This data is visualized in Grafana.

Grafana showcase

• usegalaxy.eu public server
• usegalaxy.org.au public server
• usegalaxy.org private server

If you see a dashboard you can export its configuration and put it on your Grafana with your data. Copy away!

Speaker Notes

• We have several public Grafana servers.
• If you like any of our graphs, you can copy them.

Speaker Notes

• We have built numerous dashboards for monitoring Galaxy.
• These include scripts and playbooks and configuration for everything.
• Here is EU’s galaxy dashboard showing active users, running and unscheduled jobs, etc.

Speaker Notes

• However sometimes we notice something going wrong with our infrastructure.
• We use the node detail dashboard to begin our investigation.
• It gives us a very fast overview of the server.
• This can help efficiently pinpoint isuses.

Speaker Notes

• We also monitor the database heavily.
• All of this monitoring is built into telegraf.
• We need to be able to correlate latency with autovacuums or contention.
• We monitor table size changes to check for anomalies.

Speaker Notes

• Our staff often needs to report numbers for their grants.
• We produced this user statistics dashboard to help them.
• Now they can answer their own questions, and make their own graphs, without admin help.

Speaker Notes

• We don’t just monitor Galaxy though.
• We also monitor CVMFS, and the availability of repositories in each server.
• This can give a good view of which repositories are replicated.

Key Points

• Telegraf provides an easy solution to monitor servers
• Galaxy can send metrics to Telegraf
• Telegraf can run arbitrary commands like `gxadmin`, which provides influx formatted output
• InfluxDB can collect metrics from Telegraf
• Use Grafana to visualise these metrics, and monitor their values

Thank you!