Cluster Sharding concepts

The ShardRegion actor is started on each node in the cluster, or group of nodes tagged with a specific role. The ShardRegion is created with two application specific functions to extract the entity identifier and the shard identifier from incoming messages. A Shard is a group of entities that will be managed together. For the first message in a specific shard the ShardRegion requests the location of the shard from a central coordinator, the ShardCoordinator.

The ShardCoordinator decides which ShardRegion shall own the Shard and informs that ShardRegion. The region will confirm this request and create the Shard supervisor as a child actor. The individual Entities will then be created when needed by the Shard actor. Incoming messages thus travel via the ShardRegion and the Shard to the target Entity.

If the shard home is another ShardRegion instance messages will be forwarded to that ShardRegion instance instead. While resolving the location of a shard incoming messages for that shard are buffered and later delivered when the shard home is known. Subsequent messages to the resolved shard can be delivered to the target destination immediately without involving the ShardCoordinator.

Scenarios

Once a Shard location is known ShardRegions send messages directly. Here are the scenarios for getting to this state. In the scenarios the following notation is used:

  • SC - ShardCoordinator
  • M# - Message 1, 2, 3, etc
  • SR# - ShardRegion 1, 2 3, etc
  • S# - Shard 1 2 3, etc
  • E# - Entity 1 2 3, etc. An entity refers to an Actor managed by Cluster Sharding.

Where # is a number to distinguish between instances as there are multiple in the Cluster.

Scenario 1: Message to an unknown shard that belongs to the local ShardRegion

  1. Incoming message M1 to ShardRegion instance SR1.
  2. M1 is mapped to shard S1. SR1 doesn’t know about S1, so it asks the SC for the location of S1.
  3. SC answers that the home of S1 is SR1.
  4. SR1 creates child actor shard S1 and forwards the message to it.
  5. S1 creates child actor for E1 and forwards the message to it.
  6. All incoming messages for S1 which arrive at SR1 can be handled by SR1 without SC.

Scenario 2: Message to an unknown shard that belongs to a remote ShardRegion

  1. Incoming message M2 to ShardRegion instance SR1.
  2. M2 is mapped to S2. SR1 doesn’t know about S2, so it asks SC for the location of S2.
  3. SC answers that the home of S2 is SR2.
  4. SR1 sends buffered messages for S2 to SR2.
  5. All incoming messages for S2 which arrive at SR1 can be handled by SR1 without SC. It forwards messages to SR2.
  6. SR2 receives message for S2, ask SC, which answers that the home of S2 is SR2, and we are in Scenario 1 (but for SR2).

Shard location

To make sure that at most one instance of a specific entity actor is running somewhere in the cluster it is important that all nodes have the same view of where the shards are located. Therefore the shard allocation decisions are taken by the central ShardCoordinator, which is running as a cluster singleton, i.e. one instance on the oldest member among all cluster nodes or a group of nodes tagged with a specific role.

The logic that decides where a shard is to be located is defined in a pluggable shard allocation strategy.

Shard rebalancing

To be able to use newly added members in the cluster the coordinator facilitates rebalancing of shards, i.e. migrate entities from one node to another. In the rebalance process the coordinator first notifies all ShardRegion actors that a handoff for a shard has started. That means they will start buffering incoming messages for that shard, in the same way as if the shard location is unknown. During the rebalance process the coordinator will not answer any requests for the location of shards that are being rebalanced, i.e. local buffering will continue until the handoff is completed. The ShardRegion responsible for the rebalanced shard will stop all entities in that shard by sending the specified stopMessage (default PoisonPill) to them. When all entities have been terminated the ShardRegion owning the entities will acknowledge the handoff as completed to the coordinator. Thereafter the coordinator will reply to requests for the location of the shard, thereby allocating a new home for the shard, and then buffered messages in the ShardRegion actors are delivered to the new location. This means that the state of the entities are not transferred or migrated. If the state of the entities are of importance it should be persistent (durable), e.g. with Persistence (or see Classic Persistence), so that it can be recovered at the new location.

The logic that decides which shards to rebalance is defined in a pluggable shard allocation strategy. The default implementation ShardCoordinator.LeastShardAllocationStrategy picks shards for handoff from the ShardRegion with most number of previously allocated shards. They will then be allocated to the ShardRegion with least number of previously allocated shards, i.e. new members in the cluster.

For the LeastShardAllocationStrategy there is a configurable threshold (rebalance-threshold) of how large the difference must be to begin the rebalancing. The difference between number of shards in the region with most shards and the region with least shards must be greater than the rebalance-threshold for the rebalance to occur.

A rebalance-threshold of 1 gives the best distribution and therefore typically the best choice. A higher threshold means that more shards can be rebalanced at the same time instead of one-by-one. That has the advantage that the rebalance process can be quicker but has the drawback that the the number of shards (and therefore load) between different nodes may be significantly different.

ShardCoordinator state

The state of shard locations in the ShardCoordinator is persistent (durable) with Distributed Data (or see Classic Distributed Data) to survive failures.

When a crashed or unreachable coordinator node has been removed (via down) from the cluster a new ShardCoordinator singleton actor will take over and the state is recovered. During such a failure period shards with a known location are still available, while messages for new (unknown) shards are buffered until the new ShardCoordinator becomes available.

Message ordering

As long as a sender uses the same ShardRegion actor to deliver messages to an entity actor the order of the messages is preserved. As long as the buffer limit is not reached messages are delivered on a best effort basis, with at-most once delivery semantics, in the same way as ordinary message sending.

AtLeastOnceDelivery

Reliable end-to-end messaging, with at-least-once semantics can be added by using AtLeastOnceDelivery with Classic Persistence, and see #20984 AtLeastOnceDelivery, including redelivery with a backoff.

Overhead

Some additional latency is introduced for messages targeted to new or previously unused shards due to the round-trip to the coordinator. Rebalancing of shards may also add latency. This should be considered when designing the application specific shard resolution, e.g. to avoid too fine grained shards. Once a shard’s location is known the only overhead is sending a message via the ShardRegion rather than directly.

Found an error in this documentation? The source code for this page can be found here. Please feel free to edit and contribute a pull request.