← sources

Docker Swarm in production: Are databases a complete No-Go, or am I doing it wrong?

source: reddit (reddit:docker)published 2026-09-28 13:02ingested 2026-09-29 02:42

retained text

Hey folks, quick question for the Swarm veterans out there. How are you running databases (Postgres, Redis, etc.) in production on Swarm?

Swarm is amazing for stateless apps, but persistent data is a headache. NFS volumes choke under high I/O load, and pinning DBs to specific nodes via `placement.constraints` feels like we're just using Swarm as a glorified `docker run` with a Single Point of Failure.

Is the golden rule simply: *"No databases in Swarm—use RDS/Managed Cloud"*? Or has anyone successfully deployed distributed storage (GlusterFS, Longhorn, etc.) with Swarm and been happy with the performance?

Starting to feel like we're just fighting the tool and missing the K8s comfort zone here. How are you guys handling this?

document fetched 2026-09-29 00:38 · canonical https://reddit.com/r/docker/comments/1wse1jj/docker_swarm_in_production_are_databases_a