Hi Levani,
The KIP looks good!
I have two comments:
1. In the example of the ideal standby task distribution, you should
make clear that the algorithm will either choose distributions Node-1,
Node-5, Node-9 or Node-1, Node-6, Node-8, but not both.
2. Could you formulate a bit more generic section "Changes in
HighAvailabilityTaskAssignor"? I think it is enough to state that this
KIP will NOT affect the task assignor behavior specified in KIP-441, but
it will rather extend the behavior of the distribution of standby
replicas. I think there is no need to refer to actual code.
After this changes, I am +1 on starting the vote thread.
Best,
Bruno
On 08.03.21 17:32, Levani Kokhreidze wrote:
Hello all,
Bumping this thread in case there’s any other feedback around KIP-708 [1].
If not, I will start voting thread sometime this week.
Best,
Levani
[1] -
https://cwiki.apache.org/confluence/display/KAFKA/KIP-708%3A+Rack+awarness+for+Kafka+Streams
<https://cwiki.apache.org/confluence/display/KAFKA/KIP-708:+Rack+awarness+for+Kafka+Streams>
On 4. Mar 2021, at 10:36, Levani Kokhreidze <levani.co...@gmail.com> wrote:
Hi Bruno,
Thanks a lot for the feedback.
I’ve updated KIP [1] based on suggestions.
Regards,
Levani
[1] -
https://cwiki.apache.org/confluence/display/KAFKA/KIP-708%3A+Rack+awarness+for+Kafka+Streams
<https://cwiki.apache.org/confluence/display/KAFKA/KIP-708:+Rack+awarness+for+Kafka+Streams>
On 1. Mar 2021, at 22:55, Bruno Cadonna <br...@confluent.io
<mailto:br...@confluent.io>> wrote:
clientTagPrefix
