It's an operational issue where there is some load on the application that's not a part of our load tests. The consumer offset drifts away and if it's not reset, it will poll the records that it can't match, resulting in false positives.
For now, the workaround is to set the consumer group offset using the kafka sh, for each partition, but it's not the most elegant and requires manual validations and actions
It's an operational issue where there is some load on the application that's not a part of our load tests. The consumer offset drifts away and if it's not reset, it will poll the records that it can't match, resulting in false positives.
For now, the workaround is to set the consumer group offset using the kafka sh, for each partition, but it's not the most elegant and requires manual validations and actions