Some partitions occasionally fail to be consumed when a single client consumes multiple partitions.

[Ocean52-XIE]

Description

When a client consumes multiple partitions, there is a probability that some partitions cannot be consumed. Normal when one client consumes one partition.

When it's abnormal just jump partition 3

will recover when restart the client

How to reproduce Create a topic with 8 partitions and use a client to consume

Checklist

IMPORTANT: We will close issues where the checklist has not been completed.

Please provide the following information:

• [x] librdkafka version (2.5.0): <REPLACE with e.g., v0.10.5 or a git sha. NOT "latest" or "current">
• [x] Apache Kafka version: <REPLACE with e.g., 0.10.2.3>
• [x] librdkafka client configuration: <REPLACE with e.g., message.timeout.ms=123, auto.reset.offset=earliest, ..>
• [x] Operating system: <REPLACE with e.g., Centos 5 (x64)>
• [x] Provide logs (with debug=.. as necessary) from librdkafka
• [x] Provide broker log excerpts
• [x] Critical issue

[anchitj]

Can you provide debug logs?

[emasab]

+1 for debug logs as we can verify if it's related to this bug https://github.com/confluentinc/librdkafka/issues/4804