Configuration - sorokinigor/akka-persistence-hazelcast GitHub Wiki

Please, read the reference.conf for a detailed description of the plugin properties. See unit tests for configuration examples.

Hazelcast

Hazelcast should be configured via it's xml configuration. By default, hazelcast.xml is used. It can be overridden with hazelcast.config-file.

It is highly recommended to manually configure the plugin's internal maps in the Hazelcast configuration. The default names are:

hazelcast.journal.map-name=akka-persistence-journal
hazelcast.journal.highest-deleted-sequence-number-map-name = "akka-persistence-highest-deleted-sequence-number"

hazelcast.snapshot-store.map-name = "akka-persistence-snapshot-store"

Akka

To activate the plugin, you can use this snippet:

akka.persistence {
  journal.plugin = "hazelcast.journal"
  snapshot-store.plugin = "hazelcast.snapshot-store"
}

Plugin

Internally, the plugin uses a dedicated thread pool for writes. If you want to override it, use hazelcast.write-dispatcher.

By default, persistAll methods are not supported. You can override this behaviour by:

• Enabling transactions by setting hazelcast.transaction.enabled to true. See more about Hazelcast's transactions.
• Setting hazelcast.journal.fail-on-non-atomic-persist-all to false. Internally, it starts to use putAll, but it is not an atomic operation in Hazelcast, therefore, this is likely to lead to an inconsistent state. This approach is not recommended.