Hi Guys,

I am facing a strange issue. I am using KafkaSpout and BaseBasicBolt to process 
messages from the Kafka.
There are around 1600 messages in the queue at a given point. From the worker 
log it seems the messages are being picked but when I use the Zookeeper shell 
to check the offset being consumed by the topology it gives the same offset(148 
) for a long period of time (checked for 1 hour).

The average latency time being shown on the Storm UI is 10 seconds.

I would like to know after what duration the KafkaSpout would write the offset 
being consumed back to the Zookeeper. Should I move to BaseRickBolt and do an 
explicit ack so that the offset are written back to the Zookeeper.

Thanks & Regards,
Nitin Gupta

Reply via email to