Speaker
Description
SWITCH as the national research and educational network runs an ownCloud service called SWITCHdrive for all the institutions of higher education. In order to run our service in a structured and time efficient way we have adopted a DevOps approach to operations, based on tools such as ELK, Grafana, Docker, Ansible and Python.
DevOps pillars we use:
1) Accept failure as normal mode of operation: failure in our system happens, we try to deal with it gracefully.
2) Implement gradual changes: controlled rollout of features and bug fixes allow to identify potential issue source.
3) Leverage tooling and automation: in order to facilitate change and recovery and operation tasks we will look at some tooling and automation solutions
4) Measure everything: most importantly of all is the monitoring which allows us visibility to see how the system is performing and plan for the future
In this talk we will look at how this approach has worked and what challenges we face.