RocketChat Stability

[jt13890]

Is there any documented way to have a failover system for RocketChat?
Not active-active, as that's a paid feature, but a backup in case the device RC is running on loses wifi or power for a shorter period of time.
I run it in a docker container FYI.
Thanks for any help.

[VectorPeak]

The documented HA path is multiple Rocket.Chat instances behind Traefik/a load balancer, all using the same MongoDB replica set; the docs also recommend a 3-member MongoDB replica set for HA and object storage for uploads. (docs.rocket.chat) (docs.rocket.chat)

For an unpaid/active-passive-style setup, the practical approach is a warm standby, not true failover: keep a second host ready with the same Rocket.Chat Docker config and version, keep Rocket.Chat stopped there, and have it point to the same external MongoDB replica set and shared/object file storage. If the primary host dies, move DNS / reverse proxy / VIP to the standby and start the container.

The important bit is that the data cannot live only on the box that might lose Wi-Fi or power. If MongoDB and uploaded files are on that same device, a second Rocket.Chat container won’t help much; you’d be doing restore-from-backup instead of failover. For short outages, a UPS + wired network is usually simpler and more reliable than trying to fake HA around a single Docker host.

---

If my answer solved your problem, please mark it as answered the question — I'm here to help, and honestly I'm also collecting Galaxy Brain badges along the way 😆

[jt13890]

Okay, that makes sense. Do you think ceph would work?