[
https://issues.apache.org/jira/browse/KAFKA-7610?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16685810#comment-16685810
]
Boyang Chen commented on KAFKA-7610:
------------------------------------
[~hachikuji] So we will still accept unknown member joining the group? Because
if we do that, one edge case I could think of is that a bad consumer keeps
restarting itself which generates a lot of unknown join request. What if we
hold a different map called *newMemberIds* to contain those responded member
ids? This way we are fencing zombie registration and keep the memory of join
attempts from new members so that next time when new members join the group we
would recognize them, and do a "real join" and expand the original member list.
Saving a single id should be much more memory efficient than saving a member
metadata. Each time we finished one rebalance, just erase the *newMemberIds*
map.
We could define a new error code like UNASSIGNED_MEMBER to trigger immediate
rejoin of new members. Does this extra protection make sense?
> Detect consumer failures in initial JoinGroup
> ---------------------------------------------
>
> Key: KAFKA-7610
> URL: https://issues.apache.org/jira/browse/KAFKA-7610
> Project: Kafka
> Issue Type: Improvement
> Reporter: Jason Gustafson
> Priority: Major
>
> The session timeout and heartbeating logic in the consumer allow us to detect
> failures after a consumer joins the group. However, we have no mechanism to
> detect failures during a consumer's initial JoinGroup when its memberId is
> empty. When a client fails (e.g. due to a disconnect), the newly created
> MemberMetadata will be left in the group metadata cache. Typically when this
> happens, the client simply retries the JoinGroup. Every retry results in a
> new dangling member created and left in the group. These members are doomed
> to a session timeout when the group finally finishes the rebalance, but
> before that time, they are occupying memory. In extreme cases, when a
> rebalance is delayed (possibly due to a buggy application), this cycle can
> repeat and the cache can grow quite large.
> There are a couple options that come to mind to fix the problem:
> 1. During the initial JoinGroup, we can detect failed members when the TCP
> connection fails. This is difficult at the moment because we do not have a
> mechanism to propagate disconnects from the network layer. A potential option
> is to treat the disconnect as just another type of request and pass it to the
> handlers through the request queue.
> 2. Rather than holding the JoinGroup in purgatory for an indefinite amount of
> time, we can return earlier with the generated memberId and an error code
> (say REBALANCE_IN_PROGRESS) to indicate that retry is needed to complete the
> rebalance. The consumer can then poll for the rebalance using its assigned
> memberId. And we can detect failures through the session timeout. Obviously
> this option requires a KIP (and some more thought).
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
