Handling failure in event sourcing
Failure is a regular thing in computer science. Systems crash. Hardware stops working. There are always some out-of-control variables that we cannot predict. However, we can assume failure will happen eventually and take some actions to deal with it. Akka persistence was designed to help ease the recovery from a failure when your actors need to maintain a state and cannot simply be restarted to start from fresh. In this recipe, we will review what happens to persistent actors when something goes wrong.
Getting ready
There are no prerequisites for this recipe.
How to do it…
The following scenarios can happen when using event sourcing and persistent actors:Â
- Cannot recover events from the journal: When creating a new persistent actor, the journal might not be available. Therefore, your actor cannot check whether it needs to recover its state. At this point, Akka persistence decides that it cannot continue, executes theÂ
onRecoveryFailure
 callback, and stops the...