vamossagar12
vamossagar12
@C0urante , @showuon , @yashmayya I have created this PR as per the discussions on the ticket. Plz review whenever you get the chance. Thanks!
In this PR, similar to Luke's PR, I have added the condition to do a revoking rebalance consecutively. The only difference is that whether revocation would happen or not is...
@C0urante , i made the changes based on my understanding of your suggestions. Plz review whenever you get the chance. Thanks!
> Have a look at the non-test code, left some comments. And I agree with Chris that we should not introduce exponential backoff for 1st revocation, and should reset it...
Also, I have made the changes to use the existing delay/scheduledRebalance mechanism to delay the revocation. cc @showuon , @C0urante
Thanks @C0urante . I had one question on the proposed approach regarding delays. Since the other changes are smallish in nature, I would wait for your response on that one...
> The changes look good. Left some comments. Also, could you re-enable the test in `RebalanceSourceConnectorsIntegrationTest#testMultipleWorkersRejoining`? This fix should resolve the flaky test. Thanks @showuon . I enabled it and...
> LGTM! Thanks for the improvement! Thank you !
@C0urante I made the switch back as suggested. Plz review whenever you get the chance.
> @vamossagar12 there are failing unit tests, can you check those out and ping us after you've taken a look and fixed any issues? Oops.. Sorry about that. I didn't...