Skip to content

KAFKA-9232: Coordinator new member heartbeat completion does not work for JoinGroup v3 #7753

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
hachikuji merged 8 commits into from
Dec 23, 2019
Merged

KAFKA-9232: Coordinator new member heartbeat completion does not work for JoinGroup v3 #7753

Show file tree
Hide file tree
Changes from 7 commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
61a2acf
return true when isNew
ableegoldman Nov 27, 2019
74a9fe9
reorder calls in onCompleteJoin
ableegoldman Nov 27, 2019
8392273
Revert "reorder calls in onCompleteJoin"
ableegoldman Dec 11, 2019
151477f
typo
ableegoldman Dec 11, 2019
321fb1c
should fall to new member timeout
ableegoldman Dec 18, 2019
18d441d
add unit test
ableegoldman Dec 20, 2019
d4d9209
indentation
ableegoldman Dec 20, 2019
69477f8
refactor test
ableegoldman Dec 20, 2019
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 9 additions & 3 deletions core/src/main/scala/kafka/coordinator/group/MemberMetadata.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -86,10 +86,16 @@ private[group] class MemberMetadata(var memberId: String,
}

def shouldKeepAlive(deadlineMs: Long): Boolean = {
if (isAwaitingJoin)
!isNew || latestHeartbeat + GroupCoordinator.NewMemberJoinTimeoutMs > deadlineMs
else awaitingSyncCallback != null ||

@ableegoldman ableegoldman Nov 27, 2019

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

replaced awaitingSyncCallback != null with isAwaitingSync

if (isNew) {
// New members are expired after the static join timeout
latestHeartbeat + GroupCoordinator.NewMemberJoinTimeoutMs > deadlineMs
} else if (isAwaitingJoin || isAwaitingSync) {
// Don't remove members as long as they have a request in purgatory
true
} else {
// Otherwise check for session expiration
latestHeartbeat + sessionTimeoutMs > deadlineMs
}
}

/**
Expand Down
17 changes: 16 additions & 1 deletion core/src/test/scala/unit/kafka/coordinator/group/GroupCoordinatorTest.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -313,6 +313,21 @@ class GroupCoordinatorTest {
assertEquals(Errors.INCONSISTENT_GROUP_PROTOCOL, joinGroupResult.error)
}

@Test
def testCompleteNewMemberTimeoutHeartbeat(): Unit = {
// New member joins the group
var responseFuture = sendJoinGroup(groupId, JoinGroupRequest.UNKNOWN_MEMBER_ID, protocolType, protocols, None, DefaultSessionTimeout, DefaultRebalanceTimeout, false)

@hachikuji hachikuji Dec 20, 2019

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

nit: could be val

timer.advanceClock(GroupInitialRebalanceDelay + 1)

val joinResult = Await.result(responseFuture, Duration(DefaultRebalanceTimeout + 100, TimeUnit.MILLISECONDS))
val group = groupCoordinator.groupManager.getGroup(groupId).get
assertEquals(Errors.NONE, joinResult.error)
assertEquals(0, group.allMemberMetadata.count(_.isNew))

@hachikuji hachikuji Dec 20, 2019

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Might be worth asserting latestHeartbeat as well?


// Make sure the delayed heartbeat based on the new member timeout has been completed
assertEquals(1, groupCoordinator.heartbeatPurgatory.numDelayed)

@hachikuji hachikuji Dec 20, 2019

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If we wanted to go a tad further, we could advance the clock by the session timeout and assert the member gets kicked out. That ensures that the heartbeat left behind has the right timeout.

}

@Test
def testNewMemberJoinExpiration(): Unit = {
// This tests new member expiration during a protracted rebalance. We first create a
Expand Down Expand Up @@ -1882,7 +1897,7 @@ class GroupCoordinatorTest {

val nextGenerationId = joinResult.generationId

// with no leader SyncGroup, the follower's request should failure with an error indicating
// with no leader SyncGroup, the follower's request should fail with an error indicating
// that it should rejoin
EasyMock.reset(replicaManager)
val followerSyncFuture = sendSyncGroupFollower(groupId, nextGenerationId, otherJoinResult.memberId, None)
Expand Down