Description
Understanding how a site behaves outside of the developer's machine is extremely important. Gathering metrics and events in a production system lead to a better knowledge of pain points and eases troubleshooting a lot when issues arise. This talk will explain how to setup a rock-solid visibility stack for your apps, how to feed data into it and how to extract meaningful information from that data.
We will start with the basics of exception handling and continue with the meat of this talk consisting in two parts:
1 - Events (i.e., mostly logging calls)
An introduction to logging in large production deployments:
- centralized logging with logstash and elasticsearch
- the case for structured logging
- rich dashboards with Kibana
2 - Metrics
An opinionated take on monitoring:
- Riemann, the universal metrics hub
- simple thresholds and alerts
- persisting metrics data to Graphite
- Riemann and Graphite dashboards