Hello
I am observing messages missing on the destination cluster. It seems to be connected to Assignment changed on workers.
Given the 2 log entries below:
# log from "worker-1"
[2022-03-15 12:56:01,768] INFO [ConsumerFetcherThread-0]: Topic partitions dump in fetcher thread: topic1-31:***-**-9467:0,topic2-2:7:0,topic3-16:720626:0,topic1-25:***-**-6345:0,topic3-19:817950:0,topic4-9:26268757:0,topic2-8:2:0,topic1-16:***-**-5728:0,topic1-13:***-**-3875:0,topic5-0:0:0,topic3-4:760100:0,topic1-10:***-**-3739:0,topci6-18:458068:0,topic10-9:0:0,topci6-15:457837:0,topic1-22:***-**-0401:0,topic3-25:871160:0,topic11-10:7249:0 (com.uber.stream.ureplicator.worker.ConsumerFetcherThread)\n"
# log from "worker-2"
[2022-03-15 13:02:41,054] INFO [ConsumerFetcherThread-0]Assignment changed, partitionMap [topic2-5, topic11-4, topic4-9, topic3-10, topic1-1, topci6-3, topic7-5, topic3-31, topic8-0, topic9-11, topci6-27, topic4-12, topic10-6, topci6-15, topic1-22, topic3-25, topic10-12, topic1-25, topic1-4, topic10-15, topic1-13, topci6-18, topic3-22, topic11-10, topic3-28, topic12-5], partitionResetOffsetMap: {topic1-25=***-**-6349, topic1-13=***-**-3875, topci6-18=458068, topic1-22=***-**-0408, topic3-25=871168, topic11-10=7249} (com.uber.stream.ureplicator.worker.ConsumerFetcherThread)\n
The topic3-25 last reported offset (on worker 1) was 871160; But on worker-2, the "Reset Offset Map" set topic3-25=871168.
The message missing on the destination cluster has the offset 871167. (this is one of many similar cases)
I am running the lastest version, from master, with federated mode disabled.
Is this a known behaviour?
Please, let me know if you need more info to shed some light here.
Thanks in advance
Hello
I am observing messages missing on the destination cluster. It seems to be connected to
Assignment changedon workers.Given the 2 log entries below:
The topic3-25 last reported offset (on worker 1) was
871160; But on worker-2, the "Reset Offset Map" settopic3-25=871168.The message missing on the destination cluster has the offset
871167. (this is one of many similar cases)I am running the lastest version, from master, with federated mode disabled.
Is this a known behaviour?
Please, let me know if you need more info to shed some light here.
Thanks in advance