Page History

Above we outlined the generalized JoinGroup protocol that the consumer will use. Next we show how we will implement consumer semantics on top of this protocol. Other use cases for the join group protocol would be implemented similarly.

The two phases of the group protocol correspond to subscription and assignment for the new consumer. Each member of the group submits their subscription as member metadata. The leader of the group collects all subscription in its JoinGroup response and sends the assignment as member state in SyncGroup. There are several advantages to having a single assignor:

1. Since the leader makes the assignment for the full group, it is the single source of truth for the metadata used in its decision making. This avoids the need to synchronize metadata among all members that is required in a multi-assignor approach. 
2. The leader of the group can enforce its own policy for controlling the rate of rebalancing. It doesn't have to rebalance after every metadata change, but can "batch" changes together to reduce the impact of metadata churn.
3. The leader is the only member that needs to receive the metadata from all members of the group. This reduces the overhead of the protocol.

The group protocol used by the consumer in the JoinGroup request corresponds to the assignment strategy that the leader will use to determine partition assignment. This allows the consumer to upgrade from one assignment strategy to another without downtime. The metadata corresponding to the assignment strategy can be strategy-specific, but generally it will include the group subscriptions for the member. The state returned to members in the SyncGroup will include the partitions assigned to that member.

For all assignment strategies, group members provide their subscriptions as an array of strings. This subscription can either be a list of topics or regular expressions (TODO: do we need distinguisher field to tell the difference? how about regex compatibility?). Partition assignments are provided in the SyncGroup response as an array of topics and partitions. The protocol supports custom data in both the subscription and assignment as a generic array of bytes to allow for custom assignor implementations. For example, a rack-aware assignor will generally need to propagate the rackId of each member to the leader in its subscription so that it can take it into account for assignment.  

ProtocolType => "consumer"
 
GroupProtocol  => AssignmentStrategy
  AssignmentStrategy => String
 
MemberMetadata => Version Subscription AssignmentStrategies
  Version      => int16
  Subscription => Topics UserData
    Topics     => [String]
    UserData     => Bytes
    
MemberState => Version Assignment
  Version           => int16
  Assignment        => TopicPartitions UserData
    TopicPartitions => [Topic Partitions]
      Topic      => String
      Partitions => [int32]
    UserData     => Bytes

Protocol: Briefly, this is how the protocol works for the consumer. 

1. Members JoinGroup with their respective subscriptions.
2. The leader collects member subscriptions from its JoinGroup response and performs the group assignment.  
3. All members (including the leader) send SyncGroup to find their assignment.
4. Once created, there are two cases which can trigger reassignment:
   1. Topic metadata changes which have no impact on subscriptions cause resync. The leader computes the new assignment and sends SyncGroup.
   2. Membership or subscription changes cause rejoin. 

Rolling Upgrades: To support rolling upgrades without downtime, there are two cases to consider: 

1. Changes affecting subscription: the protocol directly supports differing subscriptions, so there is no need for special handling. Members will only be assigned partitions compatible with their subscription.
2. Assignment strategy changes: to support a change to the assignment strategy, new versions must enable support both for the old assignment strategy and the new one. The coordinator will choose the old assignment strategy until all members have been updated. Then it will choose the new strategy. This preference is implicit in the order of the strategies in the JoinGroup request.

Handling Coordinator Failures: This proposal largely shares the coordinator failure cases and recovery mechanism from the initial protocol documented in Kafka 0.9 Consumer Rewrite Design. The recovery process depends on whether group state is persisted (e.g. in Zookeeper). With no persistence, then group members will generally have to rejoin the group when the new coordinator becomes active. Below, we assume persistence and show how failures are treated at the various stages of the protocol. 

1. All members send JoinGroup. 
2. Generation metadata is persisted.
3. All members send SyncGroup.
4. Sync metadata is persisted.

Coordinator failures at these steps are handled in the following ways:

1. If the coordinator fails before all members have joined the group or before group metadata has been persisted, then all members will resend their JoinGroup requests once the new coordinator is stable.
2. The coordinator may fail after group metadata has been persisted, but before all members of the group have received a response to their JoinGroup requests. The problem here is that once the new coordinator is stable, the leader may try to immediately synchronize while other members are still trying to join. However, when the coordinator receives a JoinGroup request from any member, it must abort any active synchronization and force all members to rejoin.
3. If the coordinator fails before a pending synchronization has been persisted, then all members will re-initiate the SyncGroup once the new coordinator is ready. 
4. If the coordinator fails after the metadata has been persisted, but before all members have received the SyncGroup response, then those members will initiate SyncGroup upon failover. Assuming synchronized state is persisted, then the coordinator can return that member's state immediately without forcing other members to resync. If it is not persisted, then a full group resync is required.

Other Interesting Cases:

• Leader Failures: The leader of each group is responsible for initiating group synchronization when topic metadata changes. A leader failure is detected by the coordinator through the expiration of its session timeout. The coordinator will respond by forcing all members to rejoin, which will allow a new leader to be elected.
• Assignment Failure: As mentioned above, there are several ways that the synchronization/assignment phase can fail. Generally, they are handled by having group members rejoin the group. The most interesting case is when the leader encounters an unrecoverable error when it computes the group's assignment. This could happen, for example, if group members don't agree on the assignment strategy to use. In this case, the assignment failure is forwarded to the broker which can then propagate it to awaiting members.
• Subscription Change: If a member changes its subscription, then it must force the group to be recreated by sending a JoinGroup request to the coordinator. This will cause the coordinator to reply to the other member's heartbeats with an error indicating that rejoin is needed, which will cause them to also send JoinGroup requests.
• Topic Metadata Change: The leader is responsible for detecting topic metadata changes which affect the group's subscription. When it finds a change, it can immediately compute the new assignment and initiate a SyncGroup with the coordinator.

One of the major concerns in this protocol is the size of the join group response. Since each member's metadata is included in the responses for all members, the total amount of data which the coordinator must forward in the join group responses increases quadratically with the size of the group. For example, with a per-member metadata size of 100KB, in a group of 100 members, each join group response would contain 10MB of data, which means that the coordinator would have to transfer 1GB total on every rebalance. It is therefore important to keep the size of the metadata fairly small. Even with smaller metadata size, the group can only grow so large before this becomes a concern again. However, we argue that the protocol is already unsuited to such large groups since it does not have any mechanism to cope with churn. Every time there is a membership change in the group, all members must synchronize to form next generation. If this happens often enough, as is possible with larger groups, then progress is severely restricted.

Consumer Embedded Protocol

Above we outlined the generalized JoinGroup protocol that the consumer would leverage. Next we show how we intend to implement consumer semantics on top of this protocol. Other use cases for the join group protocol would be implemented similarly. The two items that must be defined to use the join group protocol are the format of the protocol versions and the format of the protocol metadata.

ProtocolType => "consumer"
 
Protocol => AssignmentStrategy
  AssignmentStrategy => String
 
ProtocolMetadata => Subscription
  Subscription                 => Topics TopicPattern MetadataHash
    Topics                     => [String]
    TopicPattern               => String
    MetadataHash               => bytes

Subscriptions: To support differing subscriptions within the group, each member must include its own subscription in the protocol metadata. These subscriptions are forwarded to all members of the group who can then independently compute their assignment. Subscriptions can be specified either as a list of topics or as a regular expression. The latter can provide a more compact representation when subscribing to a large number of topics (e.g. if using mirror maker to replicate all the topics in a cluster).

Metadata: The metadata hash is included to ensure that each consumer has the same view of the topic metadata. A disagreement could cause an inconsistent assignment, so upon joining the group, each member checks the metadata hash of all other members to make sure they are consistent. It covers the full list of topics in the subscription set and their respective partition counts. If a regex subscription is used, then the hash covers all the topics in the cluster. If there is any disagreement on the number of partitions (e.g. due to stale metadata), then the hashes will compute differently and the consumers will refetch metadata and rejoin the group.

One potential concern in this protocol is whether a sustained disagreement might lead to continual rebalancing. This could be possible if two brokers disagree on the topic metadata for an extended period of time. While metadata should eventually converge, this issue can be resolved by having the consumers fetch their metadata from the coordinator, ensuring that they each see the same view. However, it would still be possible to have metadata disagreement if the metadata itself is changing at a very high rate.

It is worth mentioning that there is a class of assignment strategies which do not depend on consistent metadata among the consumers. For example, in a consistent hashing approach, each partition would be deterministically mapped to one of the group members. Even if two members see a different partition count for a topic, there would be no disagreement over which consumer owns each partition. The tradeoff is generally sub-optimal load balancing of partitions across consumers.

Note that the format of the metadata is an attribute of the assignment strategy. This makes it possible for different strategies to support different metadata formats. For rack-aware assignment, the metadata would also include the rack of each consumer, and the metadata hash would have to cover the leader of each partition since that governs where fetches will be sent to and the whole point of rack-aware assignment is to fetch from brokers on the same rack as the consumer. In general, any information that is used in decision making must somehow be included in the metadata.

Rolling Upgrades (Consumer)

Support for rolling upgrades is enabled through the protocol list in the join group request. For the consumer, this makes it possible to upgrade or change the assignment strategy used by the group without downtime. Note that the protocol does not distinguish between upgrades to the assignment strategy metadata and upgrades to the assignment strategy algorithm. The coordinator just looks for a protocol and version which are supported by all members of the group. It is up to the assignment strategy implementations to implement their own versioning. The following example provides an example of how an upgrade would work in practice.

Phase 1: The group consists of consumers A, B, and C each supporting version 0 of the round-robin assignment strategy.
A: (round-robin, 0)
B: (round-robin, 0)
C: (round-robin, 0)
GroupProtocol: (round-robin, 0)
 
Phase 2: A is upgraded. It now supports version 1 of the round-robin strategy in addition to version 0.
A: (round-robin, 1), (round-robin, 0)
B: (round-robin, 0)
C: (round-robin, 0)
GroupProtocol: (round-robin, 0)
 
Phase 3: B is similarly upgraded. The coordinator still chooses (round-robin, 0) since it must choose a version supported by all members.
A: (round-robin, 1), (round-robin, 0)
B: (round-robin, 1), (round-robin, 0)
C: (round-robin, 0)
GroupProtocol: (round-robin, 0)

Phase 4: C is similarly upgraded, which allows the coordinator to update the group's protocol.
A: (round-robin, 1), (round-robin, 0)
B: (round-robin, 1), (round-robin, 0)
C: (round-robin, 1), (round-robin, 0)
GroupProtocol: (round-robin, 1)

This example highlights a few interesting points. First, since the consumers have no awareness of the protocols supported by the other group members, they don't know when the update has completed and continue sending the metadata for both versions. This is not a major problem since the coordinator will only forward the associated metadata from the strategy chosen by the coordinator. Nevertheless, another round of updates would be needed to remove support for the old assignment strategy in the group.

As noted previously, the order of the assignment strategies is significant. The coordinator uses it when selecting between two or more protocols which are supported by all members. For the rolling upgrade case, this just means that the upgraded protocol should be listed first, which guarantees that the coordinator will select it after all members have been updated. 

Note that the same upgrade mechanism can also be used to change to a different assignment strategy.

Scaling Groups

As mentioned previously, the need to propagate the metadata of each member to all other members puts a significant restriction on the amount of metadata that can be used in large groups. For small and medium-sized groups, this might not be a major concern, but assignment strategies must be mindful of the metadata size and set clear scaling expectations. We have considered several approaches to deal with this problem.

Reducing the Join Payload

Dropping Topic Lists: The main contribution to the size of the join group response for the consumer case is the subscription set. One option is to remove the topic list and topic pattern and instead include only a hash of that data. This would make it impossible to simultaneously support differing subscription sets within the group, but it would still be possible to support a changing subscription set in the rolling upgrade scenario, albeit in a weaker form. More concretely, the embedded protocol would look like this: 

ProtocolType => "consumer"
 
Protocol => AssignmentStrategy
  AssignmentStrategy => String
 
ProtocolMetadata => SubscriptionHash MetadataHash
  SubscriptionHash => bytes
  MetadataHash     => bytes

Just as before, the consumers would join the group with their respective subscriptions and supported assignment strategies. The coordinator would select an assignment strategy supported by all members and forward the full member metadata to all members of the group. When the members of the group received the join group responses, they would compare the subscription hashes of all the members to find the largest matching sub-group. The members who belonged to this sub-group would assign partitions based on their own subscription sets (which match those of the rest of that sub-group), while non-members would just assign themselves an empty set of partitions and go dormant until the next rebalance. 

For example, suppose that a consumer group consists of members A, B, and C. If A and B are subscribed to (foo), and C is subscribed (foo, bar), then A and B would having matching subscription hashes and would form the largest matching sub-group. Therefore A and B would assign themselves the partitions from (foo), and C would be inactive. If consumer B was then upgraded to the subscription set (foo, bar), then B and C would form the largest sub-group and A would go inactive (obviously there would need to be a tie-breaker in case there is no largest sub-group). The major limitation should also be clear in this example: during a rolling upgrade, the capacity of the cluster will be temporarily halved, which may cause lag in the group. This would have to be considered in capacity planning.

Compression: An even simpler option is to use compression to reduce the size of the topic subscriptions. Without changing the protocol, members could compress their own metadata which is opaque to the broker anyway. For maximum compression, the protocol could be modified to support compression of the entire group metadata. Since most subscription sets in a group will probably have considerable overlap (even matching exactly in most cases), this should be very effective, though it does come at the cost of some additional complexity for the protocol.

Single Client Assignor

A more dramatic alternative is to change the protocol to allow one member to provide the assignment for the entire group. One option that we have considered to enable this is to allow assignment to piggyback on the join group requests. Here is a basic outline of how this protocol would work: 

1. As before, all members send their local metadata to the coordinator in a JoinGroupRequest.
2. The coordinator chooses one member of the group and sends only to it a JoinGroupResponse which contains all members and their respective metadata. All other members continue waiting for their initial JoinGroupRequest to be satisfied.
3. The chosen member performs the assignment for the entire group and sends it to the coordinator in a new JoinGroupRequest.
4. The coordinator distributes to each member its respective assignment in a JoinGroupResponse.

Since only one member of the group receives the full group's metadata, the quadratic increase in overall message load becomes linear as the number of group members increases. The assignment that is returned to each member can be much smaller since it only contains the information that that consumer needs (in the case of the consumer, this is just the assigned partitions, similar to the current server-side assignment protocol). It also solves the problem of metadata consistency since only one member of the group does the assignment. Finally, it allows for easier assignment implementations in general since they no longer have to be deterministic. Below we try to make this idea more concrete in order to provide a comparison with the protocol above.

Protocol: Here is a sketch of how the JoinGroup request and response schemas might be changed to support this protocol:

JoinGroupRequest => GroupId ProtocolType Phase
  GroupId              => String
  ProtocolType         => String
  Phase                => Join | Sync
  Join                 => SessionTimeout MemberId MemberMetadata
    SessionTimeout     => int32
    MemberId           => String
    MemberMetadata     => bytes
  Sync                 => ErrorCode [MemberId MemberState]
    ErrorCode          => int16
    MemberId           => String
    MemberState        => bytes
 
JoinGroupResponse => ErrorCode Phase
  ErrorCode              => int16
  GroupGenerationId      => int32
  Role                   => GroupSynchronizer | GroupMembers
  GroupSynchronizer      => [MemberId MemberMetadata]
    MemberId             => String
    MemberMetadata       => bytes
  GroupMembers           => [MemberId MemberState]   
    MemberId             => String
    MemberState          => bytes

The join group request and response objects support an intermediate phase which is used to synchronize the group's state. The coordinator chooses one member, the synchronizer, and sends a response to only its join group request. That member collects the individual metadata from each member and determines their respective state. For the consumer, the synchronization phase is where assignment is done and the state is just each member's respective list of partitions.

With a single member of the group making decisions for the entire group, it is no longer necessary for the coordinator to do any of that decision making itself. This means in particular that the multiple protocol support as provided by the protocol list in the join group request above can instead be pushed into the member metadata which is opaque to the coordinator. In fact, since only one member is responsible for setting the group state, the only requirement for a successful outcome is that the metadata from each member can be decoded by the member selected for synchronization. This makes upgrades easier as long as the metadata is compatible since there's no need for group to actually agree on the assignment strategy to be used.

It is possible for the synchronization phase to fail. For example, if the synchronizer cannot parse all of the metadata, then it needs to be able to signal its failure to the coordinator. This is similar to what happens in the current protocol when all members do not agree on the assignment strategy. The coordinator receives the error code in the join group request and forwards it to all members of the group. In the case of the consumer, the error is likely propagated to the user since retrying would not resolve the error. The coordinator must also handle the case that the member fails without a response. The simplest approach in this case is to have the coordinator fail the group and let them rejoin, but it could also select another member to perform the synchronization.

There are several heuristics that may be useful for choosing the synchronizer for the group. For example, choosing the last member to join the group reduces the chance that a failed member will be chosen since the join group protocol can take as long as a full session timeout to complete in the worst case. Another strategy would be to choose the newest member of the group, since it is the one most likely to have been upgraded, though this probably couldn't be depended on. If we exposed a metadata version in the join group request, then the coordinator would be able to choose the member with the largest version, which would probably be more a more useful mechanism to facilitate upgrades.

Consumer Embedded Protocol: The consumer's embedded protocol allows assignment strategy implementations to define their own metadata structure. The only requirement is a version field which implementations can use to verify compatibility among members. For the default assignment strategies, the member metadata corresponds to the member's subscription and the member state corresponds to the partition assignment. 

// Basic consumer embedded format
MemberMetadata       => MetadataVersion Metadata
  MetadataVersion    => String
  Metadata           => bytes
 
MemberState          => bytes
 
// For the default assignment strategy implementations shipped with Kafka
Version          => "default"
Metadata         => Subscription
  Subscription   => Topics TopicPattern
    Topics       => [String]
    TopicPattern => String
 
MemberState      => Assignment
  Assignment     => [Topic Partitions]
    Topic        => String
    Partitions   => [int32]

Rolling Upgrades: The single-assignor protocol has no built-in mechanism to deal with metadata upgrades. Instead, this is delegated to the assignor implementation. For example, the assignment strategy can use JSON or Avro as the embedded format in order to ensure compatibility between different versions. In general, this requires both forward and backward compatibility of the metadata format since the coordinator can choose either new or old version in a rolling upgrade scenario. An open question is whether the protocol could be modified to always have the coordinator select the member with the latest metadata version, which would reduce the requirement to backwards compatibility only.

Note, however, that the protocol does handle upgrades not affecting the metadata in a nice way. For example, switching between the range and round-robin assignors does not require any metadata change, so a new version could be deployed simply be changing configuration. There would be no need to have each consumer express support for both assignment strategies in parallel as in the protocol above.

Summary: This protocol skirts two of the major weaknesses of the initial client-side proposal.  Since only one member has to receive all the metadata for the group, the message overhead scales linearly with the number of members in the group. It also eliminates the potential for metadata disagreement between members since agreement is no longer required. The tradeoff is a bit more complexity. In particular, the server must implement a new state for awaiting the assignment from the synchronizer. There is also a little more work on the client, but it seems fairly straightforward since the client doesn't need any additional state tracking and can simply act based on the join group response.

KafkaConsumer API

Although the protocol is slightly more complex for the KafkaConsumer implementation, most of the details are hidden from users. Below we show the basic assignment interface that would be exposed in KafkaConsumer. The partition assigner is generic to allow for custom metadata. For simple versions, the generic type would probably be Void.

TODO:

To support client-side assignment, we'd have to make the following changes:

1. Migrate existing assignment strategies from the broker to the client. Since the assignment interface is nearly the same, this should be straightforward.
2. Modify client/server for the new join group protocol. Since we're not really changing the protocol (just the information that is passed through it), this should also be straightforward.
3. Remove offset validation from the consumer coordinator. Just a couple lines to remove for this.
4. Add support for assignment versioning (if we decide we need it). Depending on what we do, may or may not be trivial.