Chia-Ping Tsai created KAFKA-12387:
--
Summary: Avoid unnecessary copy of FetchResponse data in
handleFetchRequest
Key: KAFKA-12387
URL: https://issues.apache.org/jira/browse/KAFKA-12387
Project: Kafk
I guess I don't understand how multiple tags work together to achieve rack
awareness. I realize I could go look at how Elasticseach works, but ideally
this would be more plain in the KIP.
In particular I'm not sure how the tag approach is different than appending
multiple tags together, e.g. how i
Hi Bruno,
Thanks for the feedback. I think it makes sense.
I’ve updated the KIP [1] and tried to omit implementation details around the
algorithm.
Please let me know if the latest version looks OK.
Regards,
Levani
[1]
https://cwiki.apache.org/confluence/display/KAFKA/KIP-708%3A+Rack+awarness
Sure thing Boyang,
1) it is in proposed changes. I expanded on it a bit more now.
2) done
3) and done :)
thanks for the suggestions,
walker
On Fri, Feb 26, 2021 at 3:10 PM Boyang Chen
wrote:
> Thanks Walker. Some minor comments:
>
> 1. Could you add a reference to localThreadMetadata method in
Ismael Juma created KAFKA-12386:
---
Summary: Flush unflushed segments during recovery
Key: KAFKA-12386
URL: https://issues.apache.org/jira/browse/KAFKA-12386
Project: Kafka
Issue Type: Improvemen
Chia-Ping Tsai created KAFKA-12385:
--
Summary: Remove FetchResponse#responseData
Key: KAFKA-12385
URL: https://issues.apache.org/jira/browse/KAFKA-12385
Project: Kafka
Issue Type: Bug
+1 to adding latency metrics.
To add context on why CPU, memory and GC has a bigger impact than network
in a Mirror for compressed topics without KIP-712 is: *a failing / unstable
mirror cluster will have lag perpetually spiking having much larger impact
on e2e latencies*. To explain a bit more:
