...

Discussion thread:  TBD

JIRA: https://issues.apache.org/jira/browse/

Please keep the discussion on the mailing list rather than commenting on the wiki (wiki discussions get unwieldy fast).

...

We already aim to address 1, 4 through KIP-345 by applying group.instance.id to recognize members as static throughout restarts, and avoid sending LeaveGroupRequest when the member is under static membership, thus only using session timeout to kick off expired members. For circumstance 2, it is clear that if a member has metadata update such as assignment protocol change should require another rebalance to address this change as necessary. However, we have space to improve circumstance 3, because it is not a valid condition to trigger rebalance for the most time when the leader instance is just doing a restart under static membership. It is beneficial to distinguish whether leader is rejoining for the sake of rebalance, or is rejoining just due to service restart. By specifying the join reason of the request could entirely avoid rebalance during normal consumer bounces.

FurthermoreIn addition, as we are promoting incremental rebalances such as KIP-415, later we hope to support stateful consumers such as KStream group to have new member only taking in standby task and give them time to replay the state when first joined. These new followers need to indicate a change of status when they have finished replaying the state and ready to take active tasks. If no JoinReason is specified, brokers will not be able to distinguish the joiner's purpose: whether you are requiring an incremental rebalance, or you are just joining for restart? 

In conclusion, having JoinReason to gracefully handle the problem of rebalance necessity could simplify the implementation logic by a lot, and hide enough details to brokers' perspective on whether to move the group towards PrepareRebalance.  

Public Interfaces

We will add a new enum field to the JoinGroupRequest interface, and bump the version to v6:

...

public static Schema[] schemaVersions() {
    return new Schema[] {JOIN_GROUP_RESPONSE_V0, JOIN_GROUP_RESPONSE_V1, JOIN_GROUP_RESPONSE_V2, JOIN_GROUP_RESPONSE_V3, JOIN_GROUP_RESPONSE_V4, JOIN_GROUP_REQUEST_V5, JOIN_GROUP_REQUEST_V6};
}

For In the version one, we will just have meantime, two JoinReason enums will be introduced to handle the leader rejoin case for the very first version:

public enum JoinReason {
  BLINDRESTART("blindrestart"), // Join request from a start-uprestart/newly started consumer 
  TOPIC_METADATA_CHANGE("topic_metadata_change"); // The topic metadata has changed (must be from the leader)
}

Proposed Changes

When leader consumer detects Detecting a change of topic metadata , it should is currently be the only case when it when leader consumer wants to trigger a group rebalance. We will explicitly set the JoinReason to `topic_metadata_change` so that group coordinator will proceed to rebalance stage . If there is no member protocol change and JoinReason is `Blind`, then we shall not rebalance on known member rejoin. when hit this reason. For dynamic members (members rejoin with UNKNOWN_MEMBER_ID) this should have no effect the rebalance will still trigger because we only check join reason when the member is joining with an identity. For static members, if the JoinReason is specified as `blind``RESTART`, stable group won't trigger rebalance since this indicates a restart happen on this member and nothing should be affected for the entire group.

Be aware that JoinReason takes lower priority than current rebalance logic checking such as protocol change or unknown member join. That's why dynamic member's behavior will not be affected.

Compatibility, Deprecation, and Migration Plan

• This change doesn't involve backward incompatible change. Broker will still be able to read existing join group request, and for request < v6 the JoinReason field will be interpreted as "unknown" which shouldn't confuse broker's judgement on whether to trigger rebalance.
• User needs to restart both the broker and consumer fleet to enable this new feature. The specific order doesn't matter.

Rejected Alternatives

Not applicable so far.