Like many companies we had a need to run scheduled jobs in a distributed fashion and check up on their execution. We looked at a number of solutions including Kubernetes CronJobs and each was missing key features we found important. This session will look at an open source cron replacement built in node leveraging the Kubernetes API to schedule jobs across a cluster. It will use the Kubernetes API to schedule jobs, gather logs, and watch key Kubernetes APIs to respond to state changes in the cluster.
You will learn:
- How to call Kubernetes APIs
- How to use kubectl proxies to debug outside the cluster
- How to watch for state changes in the cluster