azure - Specify container deployment target with Kubernetes, e.g. Test/Production -
based on following setup of kubernetes on microsoft azure.
i able deploy docker containers, using same configuration settings.
we have 2 categories of containers, front-end , back-end, back-end consist of high intensive processing. latter want run on large instances, whereas front-end run on small instances.
what best option separate these, gues labeling hosts. not able find in docs, or in examples.
currently, adding labels hosts , restricting pods nodes appropriate label (e.g. 'large' / 'small') best way this. example, see examples/node-selection
kubernetes has resource model which, once implemented, allow tell scheduler resources each pod needs , system ensure pod placed on node available resources. unfortunately, can tell design document on github, isn't implemented yet.