System Health Monitoring

About

We collect metrics about the various software on which OBS relies.

Telegraf fetches metrics for each software through an input plugin, then reports those metrics to an influxDB database via the influxDB output plugin. We then display those metrics in our Grafana instance.

You can login with your GitHub account and should get the Editor role.

Implementation

Our system health monitoring consists of the following software:

• RabbitMQ: Grafana dashboard from metrics fetched by the RabbitMQ input plugin for Telegraf.
• Passenger: Grafana dashboard from metrics fetched by the Passenger input plugin for Telegraf.
• Backend: Grafana dashboard
• Postfix: Grafana dashboard from metrics fetched by the Procstat input plugin for Telegraf.
• Memcached: Grafana dashboard from metrics fetched by the Memcached input plugin for Telegraf.

Telegraf is runnning on the production host. Its global configuration is /etc/telegraf/telegraf.conf and the configuration files of its input plugins are under /etc/telegraf/telegraf.d/. We also have the same setup on the host obs-test... to test things!!!