Skip to content

Conversation

wookievx
Copy link
Contributor

@wookievx wookievx commented Apr 2, 2025

This is potential solution for #1200. Allowing for opting-in for graceful revoke handling (waiting for all the streams to finish, which should imply all the commits as well)

@wookievx wookievx force-pushed the rebalance-revoke-mode-v2 branch 2 times, most recently from 75b4766 to 07cb0f9 Compare April 2, 2025 19:11
@wookievx wookievx force-pushed the rebalance-revoke-mode-v2 branch from 07cb0f9 to 8687b76 Compare April 23, 2025 20:44
This is potential solution for fd4s#1200. Allowing for opting-in for graceful revoke handling (waiting for all the streams to finish, which should imply all the commits as well)
@wookievx wookievx force-pushed the rebalance-revoke-mode-v2 branch from 8687b76 to 64130f7 Compare July 26, 2025 13:47
@wookievx
Copy link
Contributor Author

@vlovgr It would be great if you could take a look. I am open to suggestions, feel free to comment. Just so you know: we succesfuly fixed production issues with the version from this fork, (rebalance suspension seems to be working in real life scenario).

@vlovgr
Copy link
Contributor

vlovgr commented Aug 1, 2025

This looks great @wookievx. Thanks! Will merge and release in 3.9.0.

@vlovgr vlovgr merged commit 63478cc into fd4s:series/3.x Aug 1, 2025
8 checks passed
@simonpetty
Copy link

Hope you don't mind a quick question here. Why re-use the session timeout, instead of a separate value? (We have quite a high session timeout, as we're using cooperative partition assigners).

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Development

Successfully merging this pull request may close these issues.

3 participants