This article is about getting the best out of your OpenShift Projects.
Getting to know OpenShift
The first best practice is to understand the platform a bit more. The best way we can provide that know-how is through a workshop session, as it is the fastest way we can offer to get up and running with OpenShift (book us for a workshop).
Below some great resources should you want to get some knowledge on your own:
Kubernetes documentation is very clear on the matter.
Here are our recommendations for a highly-available setup:
- Set your quotas wisely.
- Ensure that your deployments use the Rolling Deploy deployment strategy. It is the default strategy. If you run into quota problems during new rollouts, you might need to set the maxSurge parameter to 0, as explained in this article.
- Set your replicas number to 3 for your deployments
- Set anti-affinity rules for your pods, so that they all run on different nodes
- Set a PodDisruptionBudget to 2