KAFKA-14500; [2/N] Rewrite GroupMetadata in Java#13663
Conversation
dajac
left a comment
dajac
left a comment
There was a problem hiding this comment.
@jeffkbkim Thanks for the patch. I made a first pass on it and left some small comments.
|
|
||
| @BeforeEach | ||
| public void initialize() { | ||
| group = new GenericGroup(new LogContext(), "groupId", Empty, Time.SYSTEM); |
There was a problem hiding this comment.
nit: I usually prefer to avoid global variable like this in tests but I leave it up to you.
There was a problem hiding this comment.
i'll keep this as is since this variable is used for all tests.
| private GenericGroupState state; | ||
|
|
||
| /** | ||
| * The timestamp of when the group transitioned |
There was a problem hiding this comment.
this looks grammatically correct
| private final Map<String, String> staticMembers = new HashMap<>(); | ||
|
|
||
| /** | ||
| * Members who have yet to (re)join the group |
There was a problem hiding this comment.
this looks grammatically correct
dajac
left a comment
dajac
left a comment
There was a problem hiding this comment.
@jeffkbkim Thanks for the update. I left some comments for consideration.
| public boolean isGenericGroup() { | ||
| return protocolType.map(type -> | ||
| type.equals(ConsumerProtocol.PROTOCOL_TYPE) | ||
| ).orElse(false); | ||
| } |
There was a problem hiding this comment.
Was this method in the old implementation? The name is a bit weird because a generic group could use different protocol types (e.g. connect). Should it be named useConsumerGroupProtocol?
There was a problem hiding this comment.
got it. this is from
def isConsumerGroup: Boolean = protocolType.contains(ConsumerProtocol.PROTOCOL_TYPE)
the naming is confusing because we both consumer and generic groups but a generic group can expect a group using the consumer protocol
| * The timestamp of when the group transitioned | ||
| * to its current state. | ||
| */ | ||
| private Optional<Long> currentStateTimestamp; |
There was a problem hiding this comment.
Is this necessary to be an Optional? As far as I see we immediately define it in the constructor and we never set it equal to something which is empty. Am I missing something obvious?
There was a problem hiding this comment.
This is a bit awkward as the existing GroupMetadata updates this field when we read the group metadata record (https://github.com/apache/kafka/blob/trunk/core/src/main/scala/kafka/coordinator/group/GroupMetadataManager.scala#L1225).
So we should expect the new group metadata manager introduced in #13639 to perform this.
There was a problem hiding this comment.
Sorry, I do not understand. This field is currently private so the only way we can set it now is either in the constructor or via a setter. Do you mean that in the near future either we will make this field public or we will add a setter which will then be called by the GroupMetadataManager to possibly set this to an empty Optional? If so, then okay, this makes sense.
4 participants
Rewrites GroupMetadata as GenericGroup that will be used with the new group coordinator. Offset related fields, classes, and methods will not be included as they will be reworked as a separate offset manager component.
Committer Checklist (excluded from commit message)