gyfora commented on PR #403: URL: https://github.com/apache/flink-kubernetes-operator/pull/403#issuecomment-1279444456
> > We should think of a way to test these in a reliable way. Testing the parameters is not enough as might not be understood by the JM as you pointed out. > > Maybe we need some e2e tests for standalone mode that test these params properly > > I agree, e2e testing seems to be the only reliable way to verify that it works. > > I've checked the `e2e-tests` folder and it looks like it contains only fairly basic tests. To properly test this change we need to: > > 1. Start a stateful application > 2. Take a savepoint > 3. Stop it > 4. Start another stateful application with a slightly different topology so `allowNonRestoredState` is required > 5. Confirm it's restored properly > > Perhaps we can skip steps 1-3 if we're ok with keeping a savepoint file as a test resource. Also, step 4 is particularly tricky: is there a standard Flink example that we can use? Otherwise, we'd need to build a new app just for testing. > > This feel like a new test file. > > @gyfora thoughts? I agree, it would be nice to have comprehensive tests that actually test the allowNonRestoredState logic. On the other hand if we simply enable that flag on existing tests at least we provide a regression test for this particular issue. Because current test would have still failed before this fix if that flag is enabled for standalone. I think we should do something simple for this ticket and open another jira for a complete test case. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: issues-unsubscr...@flink.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
