I have a strech cluster with Scylla for 2 data centers with a balancer. The principle of operation is Every_Quirum. (After reading the documentation, I came to the conclusion that all DPCs should work)
When the connection between data centers is broken: when I writing to Scylla there are errors, that the required quorum level cannot be obtained.
Is there any way to remove a fallen data center from the balancer? Is it possible to solve the problem with the Scylla tools (orchestrator) or can only be solved using disaster recovery using triggers and scenarios? There is a suspicion that the principle of scylla is architectural, has a dedicated and complete set of components on each instance...
