apoorvmittal10 commented on code in PR #18864:
URL: https://github.com/apache/kafka/pull/18864#discussion_r1957418602
##########
group-coordinator/src/main/java/org/apache/kafka/coordinator/group/assignor/SimpleAssignor.java:
##########
@@ -72,36 +77,178 @@ private GroupAssignment assignHomogenous(
if (subscribeTopicIds.isEmpty())
return new GroupAssignment(Collections.emptyMap());
- Map<Uuid, Set<Integer>> targetPartitions = computeTargetPartitions(
+ // Subscribed topic partitions for the share group.
+ List<TargetPartition> targetPartitions = computeTargetPartitions(
subscribeTopicIds, subscribedTopicDescriber);
- return new
GroupAssignment(groupSpec.memberIds().stream().collect(Collectors.toMap(
- Function.identity(), memberId -> new
MemberAssignmentImpl(targetPartitions))));
+ // The current assignment from topic partition to members.
+ Map<TargetPartition, List<String>> currentAssignment =
currentAssignment(groupSpec);
+ return newAssignmentHomogeneous(groupSpec, subscribeTopicIds,
targetPartitions, currentAssignment);
}
private GroupAssignment assignHeterogeneous(
GroupSpec groupSpec,
SubscribedTopicDescriber subscribedTopicDescriber
) {
- Map<String, MemberAssignment> members = new HashMap<>();
+ Map<String, List<TargetPartition>> memberToPartitionsSubscription =
new HashMap<>();
for (String memberId : groupSpec.memberIds()) {
MemberSubscription spec = groupSpec.memberSubscription(memberId);
if (spec.subscribedTopicIds().isEmpty())
continue;
- Map<Uuid, Set<Integer>> targetPartitions = computeTargetPartitions(
+ // Subscribed topic partitions for the share group member.
+ List<TargetPartition> targetPartitions = computeTargetPartitions(
spec.subscribedTopicIds(), subscribedTopicDescriber);
+ memberToPartitionsSubscription.put(memberId, targetPartitions);
+ }
+
+ // The current assignment from topic partition to members.
+ Map<TargetPartition, List<String>> currentAssignment =
currentAssignment(groupSpec);
+ return newAssignmentHeterogeneous(groupSpec,
memberToPartitionsSubscription, currentAssignment);
+ }
- members.put(memberId, new MemberAssignmentImpl(targetPartitions));
+ // Get the current assignment for subscribed topic partitions to share
group members.
Review Comment:
Please write the method comments as javadoc.
##########
group-coordinator/src/main/java/org/apache/kafka/coordinator/group/assignor/SimpleAssignor.java:
##########
@@ -111,12 +258,26 @@ private Map<Uuid, Set<Integer>> computeTargetPartitions(
);
}
- Set<Integer> partitions = new HashSet<>();
for (int i = 0; i < numPartitions; i++) {
- partitions.add(i);
+ targetPartitions.add(new TargetPartition(topicId, i));
}
- targetPartitions.put(topicId, partitions);
});
return targetPartitions;
}
+
+ record TargetPartition(Uuid topicId, int partition) {
Review Comment:
Why we didn't use `org.apache.kafka.server.common.TopicIdPartition`, seems
same?
##########
group-coordinator/src/main/java/org/apache/kafka/coordinator/group/assignor/SimpleAssignor.java:
##########
@@ -111,12 +292,41 @@ private Map<Uuid, Set<Integer>> computeTargetPartitions(
);
}
- Set<Integer> partitions = new HashSet<>();
for (int i = 0; i < numPartitions; i++) {
- partitions.add(i);
+ targetPartitions.add(new TargetPartition(topicId, i));
}
- targetPartitions.put(topicId, partitions);
});
return targetPartitions;
}
+
+ static class TargetPartition {
+ Uuid topicId;
+ int partition;
Review Comment:
This can be a record class.
##########
group-coordinator/src/main/java/org/apache/kafka/coordinator/group/assignor/SimpleAssignor.java:
##########
@@ -72,36 +77,178 @@ private GroupAssignment assignHomogenous(
if (subscribeTopicIds.isEmpty())
return new GroupAssignment(Collections.emptyMap());
Review Comment:
Can we please move to Map.of now?
##########
group-coordinator/src/main/java/org/apache/kafka/coordinator/group/assignor/SimpleAssignor.java:
##########
@@ -111,12 +258,26 @@ private Map<Uuid, Set<Integer>> computeTargetPartitions(
);
}
- Set<Integer> partitions = new HashSet<>();
for (int i = 0; i < numPartitions; i++) {
- partitions.add(i);
+ targetPartitions.add(new TargetPartition(topicId, i));
}
- targetPartitions.put(topicId, partitions);
});
return targetPartitions;
}
+
+ record TargetPartition(Uuid topicId, int partition) {
+
+ @Override
+ public boolean equals(Object o) {
+ if (this == o) return true;
+ if (o == null || getClass() != o.getClass()) return false;
+ TargetPartition that = (TargetPartition) o;
+ return topicId.equals(that.topicId) && partition == that.partition;
+ }
+
+ @Override
+ public int hashCode() {
+ return Objects.hash(topicId, partition);
+ }
Review Comment:
Why do we need to define them, shouldn't record class automatically provides
comparison based on data itself?
##########
group-coordinator/src/test/java/org/apache/kafka/coordinator/group/assignor/SimpleAssignorTest.java:
##########
@@ -77,6 +85,17 @@ public void testAssignWithEmptyMembers() {
);
assertEquals(Collections.emptyMap(), groupAssignment.members());
+
+ groupSpec = new GroupSpecImpl(
+ Collections.emptyMap(),
+ HETEROGENEOUS,
+ Collections.emptyMap()
Review Comment:
Map.of() now, same above and elsewhere.
##########
group-coordinator/src/main/java/org/apache/kafka/coordinator/group/assignor/SimpleAssignor.java:
##########
@@ -72,36 +77,178 @@ private GroupAssignment assignHomogenous(
if (subscribeTopicIds.isEmpty())
return new GroupAssignment(Collections.emptyMap());
- Map<Uuid, Set<Integer>> targetPartitions = computeTargetPartitions(
+ // Subscribed topic partitions for the share group.
+ List<TargetPartition> targetPartitions = computeTargetPartitions(
subscribeTopicIds, subscribedTopicDescriber);
- return new
GroupAssignment(groupSpec.memberIds().stream().collect(Collectors.toMap(
- Function.identity(), memberId -> new
MemberAssignmentImpl(targetPartitions))));
+ // The current assignment from topic partition to members.
+ Map<TargetPartition, List<String>> currentAssignment =
currentAssignment(groupSpec);
+ return newAssignmentHomogeneous(groupSpec, subscribeTopicIds,
targetPartitions, currentAssignment);
}
private GroupAssignment assignHeterogeneous(
GroupSpec groupSpec,
SubscribedTopicDescriber subscribedTopicDescriber
) {
- Map<String, MemberAssignment> members = new HashMap<>();
+ Map<String, List<TargetPartition>> memberToPartitionsSubscription =
new HashMap<>();
for (String memberId : groupSpec.memberIds()) {
MemberSubscription spec = groupSpec.memberSubscription(memberId);
if (spec.subscribedTopicIds().isEmpty())
continue;
- Map<Uuid, Set<Integer>> targetPartitions = computeTargetPartitions(
+ // Subscribed topic partitions for the share group member.
+ List<TargetPartition> targetPartitions = computeTargetPartitions(
spec.subscribedTopicIds(), subscribedTopicDescriber);
+ memberToPartitionsSubscription.put(memberId, targetPartitions);
+ }
+
+ // The current assignment from topic partition to members.
+ Map<TargetPartition, List<String>> currentAssignment =
currentAssignment(groupSpec);
+ return newAssignmentHeterogeneous(groupSpec,
memberToPartitionsSubscription, currentAssignment);
+ }
- members.put(memberId, new MemberAssignmentImpl(targetPartitions));
+ // Get the current assignment for subscribed topic partitions to share
group members.
+ private Map<TargetPartition, List<String>> currentAssignment(GroupSpec
groupSpec) {
+ Map<TargetPartition, List<String>> assignment = new HashMap<>();
+
+ for (String member : groupSpec.memberIds()) {
+ Map<Uuid, Set<Integer>> assignedTopicPartitions =
groupSpec.memberAssignment(member).partitions();
+ assignedTopicPartitions.forEach((topicId, partitions) ->
partitions.forEach(
+ partition -> assignment.computeIfAbsent(new
TargetPartition(topicId, partition), k -> new ArrayList<>()).add(member)));
}
+ return assignment;
+ }
+
+ private GroupAssignment newAssignmentHomogeneous(
+ GroupSpec groupSpec,
+ Set<Uuid> subscribeTopicIds,
+ List<TargetPartition> targetPartitions,
+ Map<TargetPartition, List<String>> currentAssignment) {
+
+ Map<TargetPartition, List<String>> newAssignment = new HashMap<>();
+ // Step 1: Hash member IDs to partitions.
+ memberHashAssignment(targetPartitions, groupSpec.memberIds(),
newAssignment);
+
+ // Step 2: Round-robin assignment for unassigned partitions which do
not have members already assigned in the current assignment.
+ Set<TargetPartition> assignedPartitions = new
HashSet<>(newAssignment.keySet());
Review Comment:
Why do we need to create new HashSet?
##########
group-coordinator/src/main/java/org/apache/kafka/coordinator/group/assignor/SimpleAssignor.java:
##########
@@ -72,36 +77,178 @@ private GroupAssignment assignHomogenous(
if (subscribeTopicIds.isEmpty())
return new GroupAssignment(Collections.emptyMap());
- Map<Uuid, Set<Integer>> targetPartitions = computeTargetPartitions(
+ // Subscribed topic partitions for the share group.
+ List<TargetPartition> targetPartitions = computeTargetPartitions(
subscribeTopicIds, subscribedTopicDescriber);
- return new
GroupAssignment(groupSpec.memberIds().stream().collect(Collectors.toMap(
- Function.identity(), memberId -> new
MemberAssignmentImpl(targetPartitions))));
+ // The current assignment from topic partition to members.
+ Map<TargetPartition, List<String>> currentAssignment =
currentAssignment(groupSpec);
+ return newAssignmentHomogeneous(groupSpec, subscribeTopicIds,
targetPartitions, currentAssignment);
}
private GroupAssignment assignHeterogeneous(
GroupSpec groupSpec,
SubscribedTopicDescriber subscribedTopicDescriber
) {
- Map<String, MemberAssignment> members = new HashMap<>();
+ Map<String, List<TargetPartition>> memberToPartitionsSubscription =
new HashMap<>();
for (String memberId : groupSpec.memberIds()) {
MemberSubscription spec = groupSpec.memberSubscription(memberId);
if (spec.subscribedTopicIds().isEmpty())
continue;
- Map<Uuid, Set<Integer>> targetPartitions = computeTargetPartitions(
+ // Subscribed topic partitions for the share group member.
+ List<TargetPartition> targetPartitions = computeTargetPartitions(
spec.subscribedTopicIds(), subscribedTopicDescriber);
+ memberToPartitionsSubscription.put(memberId, targetPartitions);
+ }
+
+ // The current assignment from topic partition to members.
+ Map<TargetPartition, List<String>> currentAssignment =
currentAssignment(groupSpec);
+ return newAssignmentHeterogeneous(groupSpec,
memberToPartitionsSubscription, currentAssignment);
+ }
- members.put(memberId, new MemberAssignmentImpl(targetPartitions));
+ // Get the current assignment for subscribed topic partitions to share
group members.
+ private Map<TargetPartition, List<String>> currentAssignment(GroupSpec
groupSpec) {
+ Map<TargetPartition, List<String>> assignment = new HashMap<>();
+
+ for (String member : groupSpec.memberIds()) {
+ Map<Uuid, Set<Integer>> assignedTopicPartitions =
groupSpec.memberAssignment(member).partitions();
+ assignedTopicPartitions.forEach((topicId, partitions) ->
partitions.forEach(
+ partition -> assignment.computeIfAbsent(new
TargetPartition(topicId, partition), k -> new ArrayList<>()).add(member)));
}
+ return assignment;
+ }
+
+ private GroupAssignment newAssignmentHomogeneous(
+ GroupSpec groupSpec,
+ Set<Uuid> subscribeTopicIds,
+ List<TargetPartition> targetPartitions,
+ Map<TargetPartition, List<String>> currentAssignment) {
+
+ Map<TargetPartition, List<String>> newAssignment = new HashMap<>();
+ // Step 1: Hash member IDs to partitions.
+ memberHashAssignment(targetPartitions, groupSpec.memberIds(),
newAssignment);
+
+ // Step 2: Round-robin assignment for unassigned partitions which do
not have members already assigned in the current assignment.
+ Set<TargetPartition> assignedPartitions = new
HashSet<>(newAssignment.keySet());
+ List<TargetPartition> unassignedPartitions = targetPartitions.stream()
+ .filter(targetPartition ->
!assignedPartitions.contains(targetPartition))
+ .filter(targetPartition ->
!currentAssignment.containsKey(targetPartition))
+ .collect(Collectors.toList());
+
+ roundRobinAssignment(groupSpec.memberIds(), unassignedPartitions,
newAssignment);
+
+ Map<String, Set<TargetPartition>> finalAssignment = new HashMap<>();
+
+ // When combining current assignment, we need to only consider the
topics in current assignment that are also being
+ // subscribed in the new assignment as well.
+ currentAssignment.forEach((targetPartition, members) -> {
+ if (subscribeTopicIds.contains(targetPartition.topicId))
+ members.forEach(member -> {
+ if (groupSpec.memberIds().contains(member))
+ finalAssignment.computeIfAbsent(member, k -> new
HashSet<>()).add(targetPartition);
+ });
+ });
+ newAssignment.forEach((targetPartition, members) ->
members.forEach(member ->
+ finalAssignment.computeIfAbsent(member, k -> new
HashSet<>()).add(targetPartition)));
+
+ return groupAssignment(finalAssignment, groupSpec.memberIds());
+ }
+
+ private GroupAssignment newAssignmentHeterogeneous(
+ GroupSpec groupSpec,
+ Map<String, List<TargetPartition>> memberToPartitionsSubscription,
+ Map<TargetPartition, List<String>> currentAssignment) {
+
+ // Exhaustive set of all subscribed topic partitions.
+ Set<TargetPartition> targetPartitions = new LinkedHashSet<>();
+
memberToPartitionsSubscription.values().forEach(targetPartitions::addAll);
+
+ // Create a map for topic to members subscription.
+ Map<Uuid, Set<String>> topicToMemberSubscription = new HashMap<>();
+ memberToPartitionsSubscription.forEach((member, partitions) ->
+ partitions.forEach(partition ->
topicToMemberSubscription.computeIfAbsent(partition.topicId(), k -> new
LinkedHashSet<>()).add(member)));
+
+ Map<TargetPartition, List<String>> newAssignment = new HashMap<>();
+ // Step 1: Hash member IDs to partitions.
+ memberToPartitionsSubscription.forEach((member, partitions) ->
+ memberHashAssignment(partitions,
Collections.singletonList(member), newAssignment));
+
+ // Step 2: Round-robin assignment for unassigned partitions which do
not have members already assigned in the current assignment.
+ Set<TargetPartition> assignedPartitions = new
LinkedHashSet<>(newAssignment.keySet());
+ Map<Uuid, List<TargetPartition>> unassignedPartitions = new
HashMap<>();
+ targetPartitions.forEach(targetPartition -> {
+ if (!assignedPartitions.contains(targetPartition) &&
!currentAssignment.containsKey(targetPartition))
+
unassignedPartitions.computeIfAbsent(targetPartition.topicId(), k -> new
ArrayList<>()).add(targetPartition);
+ });
+
+ unassignedPartitions.keySet().forEach(unassignedTopic ->
+
roundRobinAssignment(topicToMemberSubscription.get(unassignedTopic),
unassignedPartitions.get(unassignedTopic), newAssignment));
+
+ // Step 3: We combine current assignment and new assignment.
+ Map<String, Set<TargetPartition>> finalAssignment = new HashMap<>();
+
+ // When combining current assignment, we need to only consider the
member topic subscription in current assignment
+ // which is being subscribed in the new assignment as well.
+ currentAssignment.forEach((targetPartition, members) ->
members.forEach(member -> {
+ if
(topicToMemberSubscription.getOrDefault(targetPartition.topicId(),
Collections.emptySet()).contains(member))
+ finalAssignment.computeIfAbsent(member, k -> new
HashSet<>()).add(targetPartition);
+ }));
+ newAssignment.forEach((targetPartition, members) ->
members.forEach(member ->
+ finalAssignment.computeIfAbsent(member, k -> new
HashSet<>()).add(targetPartition)));
+
+ return groupAssignment(finalAssignment, groupSpec.memberIds());
+ }
+
+ private GroupAssignment groupAssignment(
+ Map<String, Set<TargetPartition>> assignmentByMember,
+ Collection<String> allGroupMembers) {
+ Map<String, MemberAssignment> members = new HashMap<>();
+ for (Map.Entry<String, Set<TargetPartition>> entry :
assignmentByMember.entrySet()) {
+ Map<Uuid, Set<Integer>> targetPartitions = new HashMap<>();
+ entry.getValue().forEach(targetPartition ->
targetPartitions.computeIfAbsent(targetPartition.topicId(), k -> new
HashSet<>()).add(targetPartition.partition()));
+ members.put(entry.getKey(), new
MemberAssignmentImpl(targetPartitions));
+ }
+ allGroupMembers.forEach(member -> {
+ if (!members.containsKey(member))
+ members.put(member, new MemberAssignmentImpl(new HashMap<>()));
+ });
+
return new GroupAssignment(members);
}
- private Map<Uuid, Set<Integer>> computeTargetPartitions(
+ // Visible for testing.
+ void memberHashAssignment(
+ List<TargetPartition> targetPartitions,
+ Collection<String> memberIds,
+ Map<TargetPartition, List<String>> assignment) {
Review Comment:
Can we please javadoc comments explaining why to pass the
`Map<TargetPartition, List<String>> assignment` rather this method returning
`Map<TargetPartition, List<String>> assignment`.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: jira-unsubscr...@kafka.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
