You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@kafka.apache.org by "Dimitar Dimitrov (Jira)" <ji...@apache.org> on 2023/03/20 09:26:00 UTC

[jira] [Created] (KAFKA-14821) Better handle timeouts in ListOffsets API

Dimitar Dimitrov created KAFKA-14821:
----------------------------------------

             Summary: Better handle timeouts in ListOffsets API
                 Key: KAFKA-14821
                 URL: https://issues.apache.org/jira/browse/KAFKA-14821
             Project: Kafka
          Issue Type: Improvement
          Components: clients
            Reporter: Dimitar Dimitrov
            Assignee: Dimitar Dimitrov


The ListOffsets Admin API doesn't retry failed requests for partitions due to timeouts or due to other types of retriable exceptions. This is a step back compared to the Consumer offset APIs implemented in the fetcher as the latter can do partial retries in such cases.
 * The comparison is notable as some Kafka tools (e.g. {{{}kafka-get-offsets{}}}) have moved from using the Consumer offset APIs to using the ListOffsets Admin API.

One nice way to address that seems to be to migrate the ListOffsets API to use the more modern AdminApiDriver mechanism. That should automatically provide the capability to retry requests which responses are deemed retriable.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)