Given the stack trace you've shared below I can tell that this *is not a
replication issue* but rather -- your producer is not being able to
write records into the partitions because the brokers that host them are
unavailable. Now, I know that they are indeed running so "unavailable"
here means that from the network perspective your client app is not
being able to establish a TCP connection with them.
Firewall is something you should look into, as well as SELinux, and
Docker networking (if this is running on Docker).
Thanks,
-- Ricardo
On 7/6/20 10:37 AM, s.schabb...@fluent-software.de wrote:
Hi there,
I just have a problem with my kafka brokers, maybe a firewall issue, but I
don’t know. I have got 3 Brokers at three different Servers (each with another
IP) and on the first server running zookeeper:
Server1:9092 (zookeeper:2182)
Server2:9092
Server3:9092
And I have got a topic with a replication factor of three. If I try to publish
new messages to that topic, I got the following error:
Confluent.Kafka.ProduceException`2[System.Byte[],System.String]: Local: Message
timed out
at Confluent.Kafka.Producer`2.ProduceAsync(TopicPartition topicPartition,
Message`2 message, CancellationToken cancellationToken)
at
FluentSoftware.EventSourcing.KafkaProducerEventPublisher.PublishEvents[TEvent](String
topic, IEnumerable`1 eventsToApply)
at FluentSoftware.EventSourcing.CommandBus.SendAsync[TCommand](TCommand
command)
at
Heliprinter.Mediator.Communication.OrderService.Controllers.OrderController.Post(PostOrderViewModel
newOrder) in
C:\agent\_work\15\s\Source\Services\Heliprinter.Mediator.Communication.OrderService\Controllers\OrderController.cs:line
55
and the REST API responses with an Gateway timeout. Could anyone tell me
* If this error could be an replication issue?
* How I can debug these issue?
* And where I can find replication details in the logs?
Kind regards,
Sebastian.
