Raft of little operator issues

Development Kubernetes Operator Development
1

I’ve a whole raft of little operator issues that we’ve uncovered while experimenting with it; before filing them as bugs I wanted to check that thats ok : that we’re not pushing it too far or in unexpected directions.

The high level description is that we’ve been using it to manage FDB deployed to node-local storage, which incurs some additional use cases (e.g. being able to cause process exclusions, not just pod deletion with volume reuse). But also friction on version upgrades, corner cases on configmap content changes and so on; I should be able to do a decent write up of each issue :).

If, for instance, the operator is considered still experimental, then perhaps these are considered sharp edges for now, and we should just suck it up until post 1.0; but if it is meant to be fairly reliable and robust in terms of iteration and changes at this time - then I think filing what I ran into is appropriate.

2

My sense is that the operator is fairly close to 1.0, so I’d encourage you to file any issues that you run into so we can make sure it’s really stable by the time we get to 1.0.

3

This is also a timely reminder that we should create a 1.0 milestone or some other way of tracking issues for 1.0 so that we can more clearly and openly discuss the expected scope and the remaining work.

1 Like