Page History

...

The server side assignor is pluggable and the client can choose the one that it wants to use by providing its name in the heartbeat request. If the selected assignor does not exist, the group coordinator will reject the heartbeat with an UNSUPPORTED_ASSIGNOR error. The list of supported assignors will be configured in the broker configuration.

...

• The basic validations are applied. The request is rejected with the appropriate error if the validation fails.
  • The group must exist.
  • The member must exist or be created.
  • The embedded protocol must be on version >= 3.
  • The generation id must match the current member epoch. The generation id is provided in the embedded protocol.
• The group state is updated if needed.
  • The group epoch is incremented if a new assignment is required.
  • The Topics provided in the embedded consumer protocol are used to update the SubscribedTopicNames.
  • The Protocols are converted to Assignors where the UserData become the metadata and the version is set to -1.
• The current member assignment is updated if needed:
  • If the member has revoked all its partitions or the required partitions, the member can transition to its next epoch. The current target assignment become the current assignment. The group coordinator replies with the new member epoch as the generation id.
  • If the member has to revoke partitions, the group coordinator replies with the current member epoch as the generation id.
• The member transitions to CompletingRebalance state.

...

1. Lookups the group or creates it.
2. Creates the member should the member epoch be zero or checks whether it exists. If it does not exist, UNKNOWN_MEMBER_ID is returned.
3. Checks wether the member epoch matches the member epoch if in its current assignment. FENCED_MEMBER_EPOCH is returned otherwise. The member is also removed from the group.
   • There is an edge case here. When the group coordinator transitions a member to its target epoch, the heartbeat response with the new member epoch may be lost. In this case, the member will retry with the member epoch that he knows about and his request will be rejected with a FENCED_MEMBER_EPOCH. This is not optimal. Instead, the group coordinator could accept the request if the partitions owned by the members are a subset of the target partitions. If it is the case, it is safe to transition the member to its target epoch again.
4. Updates the members informations if any. The group epoch is incremented if there is any change.
5. Reconcile the member assignments as explained earlier in this document. 

...

When the group coordinator handle a ConsumerGroupPrepareAssignmentRequest ConsumerGroupDescribeRequest request:

• Checks wether the group ids exists. If it does not, GROUP_ID_NOT_FOUND is returned.
• Looks up the groups and returns the response.

...