It might have been discussed before, but I can not find any traces of such discussion. So...
what do you people think of packaging of the cluster tests as a part of a stack? We have a good start in the form of light-weight fast tests; now they can be packaged and delivered as a part of the normal release. So, one a cluster is deployed it'd be easy to quickly run a validation of it using the same set of workloads that was used for the release itself. I think we were somewhat going in that direction with test fatjar and all, but have never gotten to the point where it was actually usable for the users, not just stack developers... Thoughts? -- Cos
