I\'m going to make my first application based on Neo4j Embedded database public and ran into a serious blocking issue -
This my previous question with database deleting
Neo4j Enterprise allows you to run multiple Neo4j instances in a cluster providing you high availability and scale out for reads. To survive n
failures at the same time you need 2n+1
instances.
In case of a unclean shutdown, the expected behaviour is an automatic recovering operation upon next startup (which might take some time). In the last few years I've never experienced Neo4j in a non-recoverable state after a unclean shutdown.
For embedded use cases be sure to call gdb.shutdown()
in a JVM shutdown hook.