Tommy Brunn

Results 148 comments of Tommy Brunn

With `partitionsConsumedConcurrently`, and assuming that there's no actual resource limitation that's preventing parallelization (as is the case in the example where the consumer just waits for 1 second), each partition...

Thanks for this. This is very interesting, and I think `ERR_STREAM_WRITE_AFTER_END` is an important clue. This tells us that the stream has been closed, but something is still trying to...

Could you try [running the consumer with `DEBUG` log level](https://kafka.js.org/docs/configuration#logging) and posting the logs here? If the broker is not found in the metadata, we should refresh the metadata and...

Are you consistently seeing this error or intermittently? If it's constant, that would point to you not having configured ACLs correctly, so that your producer doesn't have the right to...

The key and the partitioner should be innocent in this. Regardless of which partition a message is assigned to, the mismatch is between the broker address and the resolved IP,...

No, order is irrelevant. It's just a mapping between node ids and hostnames.

> It seems that this is only called on KafkaJSConnectionClosedError. Do you have any recommendations here? Would it be correct to remove all brokers with the specified host and port?...

Hey Helio, and thanks for taking a look at the RedHat Managed Kafka compatibility! This code is getting rather messy now from all these conditionals everywhere, leading to some confusing...

Sure, I'm just trying to find a way that we can avoid having loads of conditionals everywhere, and can generalize those needed options so that they just become different configurations...

Do you have any logs or any more information to share? It would be helpful to have access to debug logs. How do you know that your application stops consuming...