[
https://issues.apache.org/jira/browse/KAFKA-7149?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16615254#comment-16615254
]
Guozhang Wang commented on KAFKA-7149:
--------------------------------------
[~NaviBrar] Thanks for looking into this!! Could you clarify a bit more on what
do you mean by "Maybe we can take out partitionsByHost(newly TaskIdsByHost) map
from the consumers array so that it is not replicated for all the hosts and is
sent just once."
Without the PR it is a bit hard to guess the actual approach, but I'd just
boldly assume that you have changed the format of {{partitions-by-host}} from
{code}
num.partitions-by-host, [partitions-by-host] where partitions-by-host: host,
port, num.partitions, [partitions]
{code}
to
{code}
num.tasks-by-host, [tasks-by-host] where tasks-by-host: host, port, num.tasks,
[task-ids]
{code}
And you are further optimizing it to
{code}
num.hosts, [host-tasks] where host-tasks: host, port, num.tasks, [task-ids]
{code}
?
> Reduce assignment data size to improve kafka streams scalability
> ----------------------------------------------------------------
>
> Key: KAFKA-7149
> URL: https://issues.apache.org/jira/browse/KAFKA-7149
> Project: Kafka
> Issue Type: Improvement
> Components: streams
> Reporter: Ashish Surana
> Assignee: Ashish Surana
> Priority: Major
>
> We observed that when we have high number of partitions, instances or
> stream-threads, assignment-data size grows too fast and we start getting
> below RecordTooLargeException at kafka-broker.
> Workaround of this issue is commented at:
> https://issues.apache.org/jira/browse/KAFKA-6976
> Still it limits the scalability of kafka streams as moving around 100MBs of
> assignment data for each rebalancing affects performance & reliability
> (timeout exceptions starts appearing) as well. Also this limits kafka streams
> scale even with high max.message.bytes setting as data size increases pretty
> quickly with number of partitions, instances or stream-threads.
>
> Solution:
> To address this issue in our cluster, we are sending the compressed
> assignment-data. We saw assignment-data size reduced by 8X-10X. This improved
> the kafka streams scalability drastically for us and we could now run it with
> more than 8,000 partitions.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
