[Bug] Cluster: new nodes fail to start when the oldest node has been restarted

[To-om]

Request Type

Bug

Work Environment

TheHive 4.1

Problem Description

The oldest node in cluster manages singleton actors. When it is restarted, a new manager is elected but the state is lost.
The state (in particular the status of the database) must be kept if nodes are restarted without restarting all the cluster.