Multi-Node BBB failover without recreating meetings

[Hai Hoang]

Hi guys

In a multi-node setup, if a server dies mid-meeting, can we failover to another without recreating the meeting, so user dont experience too much the server die.

I'm thinking shared Postgres/Redis for all nodes. Tools like Scalelite (load balancing) or b3scale (HA/multi-tenant) seem close—do they support live handoffs?

Has anyone pulled off failover in prod? Really love for any idea Thank you :D