Persistence Plugins

Storage backends for journals and snapshot stores are pluggable in the Pekko persistence extension.

Plugins maintained within the Pekko organization are:

• pekko-persistence-cassandra (no Durable State support)
• pekko-persistence-jdbc (Durable State only supported with Postgres and H2)
• pekko-persistence-r2dbc
• pekko-persistence-dynamodb

Plugins can be selected either by “default” for all persistent actors, or “individually”, when a persistent actor defines its own set of plugins.

When a persistent actor does NOT override the journalPluginId and snapshotPluginId methods, the persistence extension will use the “default” journal, snapshot-store and durable-state plugins configured in reference.conf:

pekko.persistence.journal.plugin = ""
pekko.persistence.snapshot-store.plugin = ""
pekko.persistence.state.plugin = ""

However, these entries are provided as empty "", and require explicit user configuration via override in the user application.conf.

• For an example of a journal plugin which writes messages to LevelDB see Local LevelDB journal.
• For an example of a snapshot store plugin which writes snapshots as individual files to the local filesystem see Local snapshot store.
• The state store is relatively new, one available implementation is the pekko-persistence-jdbc-plugin.

Eager initialization of persistence plugin

By default, persistence plugins are started on-demand, as they are used. In some case, however, it might be beneficial to start a certain plugin eagerly. In order to do that, you should first add org.apache.pekko.persistence.Persistence under the pekko.extensions key. Then, specify the IDs of plugins you wish to start automatically under pekko.persistence.journal.auto-start-journals and pekko.persistence.snapshot-store.auto-start-snapshot-stores.

For example, if you want eager initialization for the leveldb journal plugin and the local snapshot store plugin, your configuration should look like this:

pekko {

  extensions = [org.apache.pekko.persistence.Persistence]

  persistence {

    journal {
      plugin = "pekko.persistence.journal.leveldb"
      auto-start-journals = ["org.apache.pekko.persistence.journal.leveldb"]
    }

    snapshot-store {
      plugin = "pekko.persistence.snapshot-store.local"
      auto-start-snapshot-stores = ["org.apache.pekko.persistence.snapshot-store.local"]
    }

  }

}

Pre-packaged plugins

The Pekko Persistence module comes with few built-in persistence plugins, but none of these are suitable for production usage in a Pekko Cluster.

Local LevelDB journal

This plugin writes events to a local LevelDB instance.

The LevelDB journal is deprecated and it is not advised to build new applications with it. As a replacement we recommend using Pekko Persistence JDBC.

The LevelDB journal plugin config entry is pekko.persistence.journal.leveldb. Enable this plugin by defining config property:

copysource# Path to the journal plugin to be used
pekko.persistence.journal.plugin = "pekko.persistence.journal.leveldb"

LevelDB based plugins will also require the following additional dependency declaration:

sbt

libraryDependencies += "org.fusesource.leveldbjni" % "leveldbjni-all" % "1.8"

Maven

<dependencies>
  <dependency>
    <groupId>org.fusesource.leveldbjni</groupId>
    <artifactId>leveldbjni-all</artifactId>
    <version>1.8</version>
  </dependency>
</dependencies>

Gradle

dependencies {
  implementation "org.fusesource.leveldbjni:leveldbjni-all:1.8"
}

The default location of LevelDB files is a directory named journal in the current working directory. This location can be changed by configuration where the specified path can be relative or absolute:

copysourcepekko.persistence.journal.leveldb.dir = "target/journal"

With this plugin, each actor system runs its own private LevelDB instance.

One peculiarity of LevelDB is that the deletion operation does not remove messages from the journal, but adds a “tombstone” for each deleted message instead. In the case of heavy journal usage, especially one including frequent deletes, this may be an issue as users may find themselves dealing with continuously increasing journal sizes. To this end, LevelDB offers a special journal compaction function that is exposed via the following configuration:

copysource# Number of deleted messages per persistence id that will trigger journal compaction
pekko.persistence.journal.leveldb.compaction-intervals {
  persistence-id-1 = 100
  persistence-id-2 = 200
  # ...
  persistence-id-N = 1000
  # use wildcards to match unspecified persistence ids, if any
  "*" = 250
}

Shared LevelDB journal

The LevelDB journal is deprecated and will be removed from a future Pekko version, it is not advised to build new applications with it. For testing in a multi node environment the “inmem” journal together with the proxy plugin can be used, but the actual journal used in production of applications is also a good choice.

A shared LevelDB instance is started by instantiating the SharedLeveldbStore actor.

Scala

copysourceimport org.apache.pekko.persistence.journal.leveldb.SharedLeveldbStore

val store = system.actorOf(Props[SharedLeveldbStore](), "store")

Java

copysourcefinal ActorRef store = system.actorOf(Props.create(SharedLeveldbStore.class), "store");

By default, the shared instance writes journaled messages to a local directory named journal in the current working directory. The storage location can be changed by configuration:

copysourcepekko.persistence.journal.leveldb-shared.store.dir = "target/shared"

Actor systems that use a shared LevelDB store must activate the pekko.persistence.journal.leveldb-shared plugin.

copysourcepekko.persistence.journal.plugin = "pekko.persistence.journal.leveldb-shared"

This plugin must be initialized by injecting the (remote) SharedLeveldbStore actor reference. Injection is done by calling the SharedLeveldbJournal.setStore method with the actor reference as argument.

Scala

copysourcetrait SharedStoreUsage extends Actor {
  override def preStart(): Unit = {
    context.actorSelection("pekko://example@127.0.0.1:7355/user/store") ! Identify(1)
  }

  def receive = {
    case ActorIdentity(1, Some(store)) =>
      SharedLeveldbJournal.setStore(store, context.system)
  }
}

Java

copysourceclass SharedStorageUsage extends AbstractActor {
  @Override
  public void preStart() throws Exception {
    String path = "pekko://example@127.0.0.1:7355/user/store";
    ActorSelection selection = getContext().actorSelection(path);
    selection.tell(new Identify(1), getSelf());
  }

  @Override
  public Receive createReceive() {
    return receiveBuilder()
        .match(
            ActorIdentity.class,
            ai -> {
              if (ai.correlationId().equals(1)) {
                Optional<ActorRef> store = ai.getActorRef();
                if (store.isPresent()) {
                  SharedLeveldbJournal.setStore(store.get(), getContext().getSystem());
                } else {
                  throw new RuntimeException("Couldn't identify store");
                }
              }
            })
        .build();
  }
}

Internal journal commands (sent by persistent actors) are buffered until injection completes. Injection is idempotent i.e. only the first injection is used.

Local snapshot store

This plugin writes snapshot files to the local filesystem.

The local snapshot store plugin config entry is pekko.persistence.snapshot-store.local. Enable this plugin by defining config property:

copysource# Path to the snapshot store plugin to be used
pekko.persistence.snapshot-store.plugin = "pekko.persistence.snapshot-store.local"

The default storage location is a directory named snapshots in the current working directory. This can be changed by configuration where the specified path can be relative or absolute:

copysourcepekko.persistence.snapshot-store.local.dir = "target/snapshots"

Note that it is not mandatory to specify a snapshot store plugin. If you don’t use snapshots you don’t have to configure it.

Persistence Plugin Proxy

For testing purposes a persistence plugin proxy allows sharing of a journal and snapshot store on a single node across multiple actor systems (on the same or on different nodes). This, for example, allows persistent actors to failover to a backup node and continue using the shared journal instance from the backup node. The proxy works by forwarding all the journal/snapshot store messages to a single, shared, persistence plugin instance, and therefore supports any use case supported by the proxied plugin.

The journal and snapshot store proxies are controlled via the pekko.persistence.journal.proxy and pekko.persistence.snapshot-store.proxy configuration entries, respectively. Set the target-journal-plugin or target-snapshot-store-plugin keys to the underlying plugin you wish to use (for example: pekko.persistence.journal.inmem). The start-target-journal and start-target-snapshot-store keys should be set to on in exactly one actor system - this is the system that will instantiate the shared persistence plugin. Next, the proxy needs to be told how to find the shared plugin. This can be done by setting the target-journal-address and target-snapshot-store-address configuration keys, or programmatically by calling the PersistencePluginProxy.setTargetLocation method.