Github user apiwoni commented on the issue:
https://github.com/apache/storm/pull/1679
We are on Apache Store 1.0.4 and I''m trying to convince people to upgrade
because I suspect we are experiencing this bug.
Here's what I see based on nextTuple cycle:
LOG: Topic partitions with entries ready to be retried [[TOPIC-0]]
LOG: topic-partition [TOPIC-0] has non-continuous offset [219355]. It will
be processed in a subsequent batch.
org.apache.kafka.clients.consumer.KafkaConsumer#seekToEnd[ where current
committed offset in Kafka = 219351]
Polled [0] records from Kafka. [6253] uncommitted offsets across all topic
partitions
I would expect org.apache.storm.kafka.spout.KafkaSpout#pollKafkaBroker
return few records after seekToEnd?
---
