Skip to content
This repository has been archived by the owner on Mar 24, 2021. It is now read-only.

degraded consumer operation when partitions have unrecoverable errors #810

Open
emmettbutler opened this issue May 26, 2018 · 1 comment

Comments

@emmettbutler
Copy link
Contributor

Consider a mode of operation for the consumer in which some partitions being inaccessible (possibly due to GroupLoadInProgressCode errors) does not cause the consumer to crash. fetch_offsets in particular doesn't return successfully unless all partitions provide offset information without error. It's conceivable that the consumer could consume only from the partitions it can successfully fetch offset information for, while ignoring the others.

@emmettbutler
Copy link
Contributor Author

Anecdotally, the Java client has the same issue as pykafka in this respect.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Projects
None yet
Development

No branches or pull requests

1 participant