wellingr / akka-persistence-jdbc

Asynchronously writes journal and snapshot entries to configured JDBC databases so that Akka Actors can recover state

GitHub

akka-persistence-jdbc

Download License

akka-persistence-jdbc writes journal and snapshot entries to a configured JDBC store. It implements the full akka-persistence-query API and is therefore very useful for implementing DDD-style application models using Akka and Scala for creating reactive applications.

Installation

Add the following to your build.sbt:

// the library is available in Bintray's JCenter
resolvers += Resolver.jcenterRepo

// akka 2.5.x
libraryDependencies += "com.github.wellingr" %% "akka-persistence-jdbc" % "3.0.0"

Contribution policy

Contributions via GitHub pull requests are gladly accepted from their original author. Along with any pull requests, please state that the contribution is your original work and that you license the work to the project under the project's open source license. Whether or not you state this explicitly, by submitting any copyrighted material via pull request, email, or other means you agree to license the material under the project's open source license and warrant that you have the legal authority to do so.

Code of Conduct

Contributors all agree to follow the W3C Code of Ethics and Professional Conduct.

If you want to take action, feel free to contact Dennis Vriend dnvriend@gmail.com. You can also contact W3C Staff as explained in W3C Procedures.

License

This source code is made available under the Apache 2.0 License. The quick summary of what this license means is available here

Configuration

The plugin relies on Slick 3.2.x to do create the SQL dialect for the database in use, therefore the following must be configured in application.conf

Configure akka-persistence:

  • instruct akka persistence to use the jdbc-journal plugin,
  • instruct akka persistence to use the jdbc-snapshot-store plugin,

Configure slick:

  • The following slick drivers are supported:
    • slick.jdbc.PostgresProfile$
    • slick.jdbc.MySQLProfile$
    • slick.jdbc.H2Profile$
    • slick.jdbc.OracleProfile$

Database Schema

Configuration

DataSource lookup by JNDI name

The plugin uses slick as the database access library. Slick supports jndi for looking up DataSources.

To enable the JNDI lookup, you must add the following to your application.conf:

jdbc-journal {
  slick {
    driver = "slick.jdbc.PostgresDriver$"
    jndiName = "java:jboss/datasources/PostgresDS"   
  }
}

How to get the ReadJournal using Scala

The ReadJournal is retrieved via the akka.persistence.query.PersistenceQuery extension:

import akka.persistence.query.PersistenceQuery
import akka.persistence.jdbc.query.scaladsl.JdbcReadJournal

val readJournal: JdbcReadJournal = PersistenceQuery(system).readJournalFor[JdbcReadJournal](JdbcReadJournal.Identifier)

How to get the ReadJournal using Java

The ReadJournal is retrieved via the akka.persistence.query.PersistenceQuery extension:

import akka.persistence.query.PersistenceQuery
import akka.persistence.jdbc.query.javadsl.JdbcReadJournal

final JdbcReadJournal readJournal = PersistenceQuery.get(system).getReadJournalFor(JdbcReadJournal.class, JdbcReadJournal.Identifier());

Persistence Query

The plugin supports the following queries:

AllPersistenceIdsQuery and CurrentPersistenceIdsQuery

allPersistenceIds and currentPersistenceIds are used for retrieving all persistenceIds of all persistent actors.

import akka.actor.ActorSystem
import akka.stream.{Materializer, ActorMaterializer}
import akka.stream.scaladsl.Source
import akka.persistence.query.PersistenceQuery
import akka.persistence.jdbc.query.scaladsl.JdbcReadJournal

implicit val system: ActorSystem = ActorSystem()
implicit val mat: Materializer = ActorMaterializer()(system)
val readJournal: JdbcReadJournal = PersistenceQuery(system).readJournalFor[JdbcReadJournal](JdbcReadJournal.Identifier)

val willNotCompleteTheStream: Source[String, NotUsed] = readJournal.allPersistenceIds()

val willCompleteTheStream: Source[String, NotUsed] = readJournal.currentPersistenceIds()

The returned event stream is unordered and you can expect different order for multiple executions of the query.

When using the allPersistenceIds query, the stream is not completed when it reaches the end of the currently used persistenceIds, but it continues to push new persistenceIds when new persistent actors are created.

When using the currentPersistenceIds query, the stream is completed when the end of the current list of persistenceIds is reached, thus it is not a live query.

The stream is completed with failure if there is a failure in executing the query in the backend journal.

EventsByPersistenceIdQuery and CurrentEventsByPersistenceIdQuery

eventsByPersistenceId and currentEventsByPersistenceId is used for retrieving events for a specific PersistentActor identified by persistenceId.

import akka.actor.ActorSystem
import akka.stream.{Materializer, ActorMaterializer}
import akka.stream.scaladsl.Source
import akka.persistence.query.{ PersistenceQuery, EventEnvelope }
import akka.persistence.jdbc.query.scaladsl.JdbcReadJournal

implicit val system: ActorSystem = ActorSystem()
implicit val mat: Materializer = ActorMaterializer()(system)
val readJournal: JdbcReadJournal = PersistenceQuery(system).readJournalFor[JdbcReadJournal](JdbcReadJournal.Identifier)

val willNotCompleteTheStream: Source[EventEnvelope, NotUsed] = readJournal.eventsByPersistenceId("some-persistence-id", 0L, Long.MaxValue)

val willCompleteTheStream: Source[EventEnvelope, NotUsed] = readJournal.currentEventsByPersistenceId("some-persistence-id", 0L, Long.MaxValue)

You can retrieve a subset of all events by specifying fromSequenceNr and toSequenceNr or use 0L and Long.MaxValue respectively to retrieve all events. Note that the corresponding sequence number of each event is provided in the EventEnvelope, which makes it possible to resume the stream at a later point from a given sequence number.

The returned event stream is ordered by sequence number, i.e. the same order as the PersistentActor persisted the events. The same prefix of stream elements (in same order) are returned for multiple executions of the query, except for when events have been deleted.

The stream is completed with failure if there is a failure in executing the query in the backend journal.

EventsByTag and CurrentEventsByTag

eventsByTag and currentEventsByTag are used for retrieving events that were marked with a given tag, e.g. all domain events of an Aggregate Root type.

import akka.actor.ActorSystem
import akka.stream.{Materializer, ActorMaterializer}
import akka.stream.scaladsl.Source
import akka.persistence.query.{ PersistenceQuery, EventEnvelope }
import akka.persistence.jdbc.query.scaladsl.JdbcReadJournal

implicit val system: ActorSystem = ActorSystem()
implicit val mat: Materializer = ActorMaterializer()(system)
val readJournal: JdbcReadJournal = PersistenceQuery(system).readJournalFor[JdbcReadJournal](JdbcReadJournal.Identifier)

val willNotCompleteTheStream: Source[EventEnvelope, NotUsed] = readJournal.eventsByTag("apple", 0L)

val willCompleteTheStream: Source[EventEnvelope, NotUsed] = readJournal.currentEventsByTag("apple", 0L)

Tagging events

To tag events you'll need to create an Event Adapter that will wrap the event in a akka.persistence.journal.Tagged class with the given tags. The Tagged class will instruct akka-persistence-jdbc to tag the event with the given set of tags.

The persistence plugin will not store the Tagged class in the journal. It will strip the tags and payload from the Tagged class, and use the class only as an instruction to tag the event with the given tags and store the payload in the message field of the journal table.

import akka.persistence.journal.{ Tagged, WriteEventAdapter }
import com.github.dnvriend.Person.{ LastNameChanged, FirstNameChanged, PersonCreated }

class TaggingEventAdapter extends WriteEventAdapter {
  override def manifest(event: Any): String = ""

  def withTag(event: Any, tag: String) = Tagged(event, Set(tag))

  override def toJournal(event: Any): Any = event match {
    case _: PersonCreated 
      withTag(event, "person-created")
    case _: FirstNameChanged 
      withTag(event, "first-name-changed")
    case _: LastNameChanged 
      withTag(event, "last-name-changed")
    case _  event
  }
}

The EventAdapter must be registered by adding the following to the root of application.conf Please see the demo-akka-persistence-jdbc project for more information.

jdbc-journal {
  event-adapters {
    tagging = "com.github.dnvriend.TaggingEventAdapter"
  }
  event-adapter-bindings {
    "com.github.dnvriend.Person$PersonCreated" = tagging
    "com.github.dnvriend.Person$FirstNameChanged" = tagging
    "com.github.dnvriend.Person$LastNameChanged" = tagging
  }
}

You can retrieve a subset of all events by specifying offset, or use 0L to retrieve all events with a given tag. The offset corresponds to an ordered sequence number for the specific tag. Note that the corresponding offset of each event is provided in the EventEnvelope, which makes it possible to resume the stream at a later point from a given offset.

In addition to the offset the EventEnvelope also provides persistenceId and sequenceNr for each event. The sequenceNr is the sequence number for the persistent actor with the persistenceId that persisted the event. The persistenceId + sequenceNr is an unique identifier for the event.

The returned event stream contains only events that correspond to the given tag, and is ordered by the creation time of the events. The same stream elements (in same order) are returned for multiple executions of the same query. Deleted events are not deleted from the tagged event stream.

Custom DAO Implementation

The plugin supports loading a custom DAO for the journal and snapshot. You should implement a custom Data Access Object (DAO) if you wish to alter the default persistency strategy in any way, but wish to reuse all the logic that the plugin already has in place, eg. the Akka Persistence Query API. For example, the default persistency strategy that the plugin supports serializes journal and snapshot messages using a serializer of your choice and stores them as byte arrays in the database.

By means of configuration in application.conf a DAO can be configured, below the default DAOs are shown:

jdbc-journal {
  dao = "akka.persistence.jdbc.journal.dao.ByteArrayJournalDao"
}

jdbc-snapshot-store {
  dao = "akka.persistence.jdbc.snapshot.dao.ByteArraySnapshotDao"
}

jdbc-read-journal {
  dao = "akka.persistence.jdbc.query.dao.ByteArrayReadJournalDao"
}

Storing messages as byte arrays in blobs is not the only way to store information in a database. For example, you could store messages with full type information as a normal database rows, each event type having its own table. For example, implementing a Journal Log table that stores all persistenceId, sequenceNumber and event type discriminator field, and storing the event data in another table with full typing

You only have to implement two interfaces akka.persistence.jdbc.journal.dao.JournalDao and/or akka.persistence.jdbc.snapshot.dao.SnapshotDao. As these APIs are only now exposed for public use, the interfaces may change when the API needs to change for whatever reason.

For example, take a look at the following two custom DAOs:

class MyCustomJournalDao(db: Database, val profile: JdbcProfile, journalConfig: JournalConfig, serialization: Serialization)(implicit ec: ExecutionContext, mat: Materializer) extends JournalDao {
    // snip
}

class MyCustomSnapshotDao(db: JdbcBackend#Database, val profile: JdbcProfile, snapshotConfig: SnapshotConfig, serialization: Serialization)(implicit ec: ExecutionContext, val mat: Materializer) extends SnapshotDao {
    // snip
}

As you can see, the custom DAOs get a Slick database, a Slick profile, the journal or snapshot configuration, an akka.serialization.Serialization, an ExecutionContext and Materializer injected after constructed. You should register the Fully Qualified Class Name in application.conf so that the custom DAOs will be used.

For more information please review the two default implementations akka.persistence.jdbc.dao.bytea.journal.ByteArrayJournalDao and akka.persistence.jdbc.dao.bytea.snapshot.ByteArraySnapshotDao or the demo custom DAO example from the demo-akka-persistence site.

Explicitly shutting down the database connections

The plugin automatically shuts down the HikariCP connection pool only when the ActorSystem is explicitly terminated. It is advisable to register a shutdown hook to be run when the VM exits that terminates the ActorSystem:

sys.addShutdownHook(system.terminate())

Changelog

3.0.0 (TO BE RELEASED)

  • Forked from dnvriend-akka-persistence-jdbc
  • EventsByTag query treats the offset as exclusive instead of inclusive. This is a breaking change!
  • Increased major version number.
  • Fixed bug in eventsByTag and currentEventsByTag query, which could case some event to be skipped
  • Implemented batch writing for the journal which results in better thoughput.
  • Fixed a potential issue where actors could retrieve their latest sequence number too early upon restart.
  • Akka 2.5.2 -> 2.5.4
  • Slick 3.2.0 -> 3.2.1

2.4.18.2 (2017-06-09)

  • Fixed Issue #106 'JdbcReadJournal in javadsl does not support CurrentEventsByTagQuery2 and EventsByTagQuery2'

2.5.2.0 (2017-06-09)

  • Merged PR #105 aenevala Added support for event adapters on query side for 2.5.2.0, thanks!
  • Akka 2.5.1 -> 2.5.2

2.4.18.1 (2017-06-04)

  • Merged PR #103 aenevala Added support for event adapters on query side, thanks!

2.5.1.0 (2017-05-03)

  • Akka 2.5.0 -> 2.5.1

2.4.18.0 (2017-05-03)

2.5.0.0 (2017-04-13)

  • Merged PR #97 Dan Di Spaltro Update to akka-2.5.0-RC2, thanks!
  • Akka 2.5.0-RC2 -> Akka 2.5.0

2.4.17.1 (2017-02-24)

  • Slick 3.1.1 -> 3.2.0
  • Scala 2.11.8 and 2.12.1 support
  • The following slick drivers are supported:
    • slick.jdbc.PostgresProfile$
    • slick.jdbc.MySQLProfile$
    • slick.jdbc.H2Profile$
    • slick.jdbc.OracleProfile$

2.4.17.0.3.2.0-RC1

  • Slick 3.2.0-RC1 test release
  • Akka 2.4.17
  • Only availabe on Bintray's JCenter

2.4.17.0 (2016-02-12)

  • New versioning scheme; now using the version of Akka with the akka-persistence-inmemory version appended to it, starting from .0
  • Akka 2.4.16 -> 2.4.17

2.6.12-3.2.0-M2 (2016-12-21)

  • Special thanks to joseblas for PR #85: Slick 3.2.0-M1 migration, thanks!
  • Special thanks to Timothy Klim for PR #86: Upgrade slick to 3.2.0-M2, thanks!
  • Slick 3.2.0-M2 test release
  • Scala 2.11.8 and 2.12.1 build
  • Only availabe on Bintray's JCenter
  • The following slick drivers are supported:
    • slick.jdbc.PostgresProfile$
    • slick.jdbc.MySQLProfile$
    • slick.jdbc.H2Profile$
    • slick.jdbc.OracleProfile$

2.6.12 (2016-12-20)

  • Akka 2.4.14 -> 2.4.16

2.6.11 (2016-12-09)

  • Merged PR #3 Sergey Kisel - Freeslick support to use it with Oracle, thanks!

2.6.10 (2016-11-22)

  • Akka 2.4.13 -> 2.4.14

2.6.9 (2016-11-20)

  • Akka 2.4.12 -> 2.4.13

2.6.8 (2016-11-03)

  • Akka 2.4.10 -> 2.4.12
  • Fixed 'Snapshot storage BLOB handling' by Sergey Kisel, thanks!
  • Filter out events that have already been deleted.
  • Removed the non-official and never-to-be-used bulk loading interface.
  • Support for the new queries CurrentEventsByTagQuery2 and EventsByTagQuery2, please read the akka-persistence-query documentation to see what has changed.
  • The akka-persistence-jdbc plugin only supports the akka.persistence.query.NoOffset and the akka.persistence.query.Sequence offset types.
  • There is no support for the akka.persistence.query.TimeBasedUUID offset type. When used, akka-persistence-jdbc will throw an IllegalArgumentException if offered to the read-journal.

2.6.7 (2016-09-07)

  • Merged PR #75 jroper - Removed binary dependency on slick-extensions, thanks!
  • Please note, slick-extensions 3.1.0 are open source, but the license didn't change, so you cannot use it for free, you still need a Lightbend Subscription.
  • Akka 2.4.9 -> 2.4.10

2.6.6 (2016-08-22)

  • Merged PR #66 monktastic, eventsByPersistenceId should terminate when toSequenceNr is reached, thanks!

2.6.5 (2016-08-20)

  • Akka 2.4.9-RC2 -> 2.4.9

2.6.5-RC2 (2016-08-06)

  • Akka 2.4.9-RC1 -> 2.4.9-RC2

2.6.5-RC1 (2016-08-03)

  • Akka 2.4.8 -> 2.4.9-RC1

2.6.4 (2016-07-30)

  • Merged PR #62 jtysper, Fix Oracle support, thanks!

2.6.3 (2016-07-27)

2.6.2 (2016-07-26)

  • Fix for issue #60 where an immutable.Vector was trying to be matched by the serializer in TrySeq stage.

2.6.1 (2016-07-23)

  • Support for the non-official bulk loading interface akka.persistence.query.scaladsl.EventWriter added. I need this interface to load massive amounts of data, that will be processed by many actors, but initially I just want to create and store one or more events belonging to an actor, that will handle the business rules eventually. Using actors or a shard region for that matter, just gives to much actor life cycle overhead ie. too many calls to the data store. The akka.persistence.query.scaladsl.EventWriter interface is non-official and puts all responsibility of ensuring the integrity of the journal on you. This means when some strange things are happening caused by wrong loading of the data, and therefore breaking the integrity and ruleset of akka-persistence, all the responsibility on fixing it is on you, and not on the Akka team.

2.6.0 (2016-07-17)

  • Removed the deleted_to and created columns of the journal table to become compatible with akka-persistence-query spec that states that all messages should be replayed, even deleted ones
  • New schema's are available for postgres, mysql and h2
  • No need for Query Publishers with the new akka-streams API
  • Codacy code cleanup
  • There is still no support for Oracle since the addition of the ordering SERIAL column which Oracle does not support. Help to add Oracle support is appreciated.

2.5.2 (2016-07-03)

  • The eventsByTag query should now be fixed.

2.5.1 (2016-07-03)

  • There is no 2.5.1; error while releasing

2.5.0 (2016-06-29)

  • Changed the database schema to include two new columns, an ordering and a deleted column. Both fields are needed to support the akka-persistence-query API. The ordering column is needed to register the total ordering of events an is used for the offset for both *byTag queries. The deleted column is not yet used.
  • Known issue: will not work on Oracle (yet).