Skip to main content

The SnoutData stack

The stack is what a project IS: a Postgres database and the servers in front of it, behind one HTTPS door. A hosted project on SnoutData Cloud runs it, and so does a folder on your own machine with Docker. Same servers, same paths, same two keys, same client code; only the URL changes. Every part is open source.

What is in it​

Where it runs​

Hosted (SnoutData Cloud)Your own Docker
Postgres 18, extensions, SnoutTimeyesyes
Data API, auth, storage, realtime, functions, pushyesyes
Socket.IOyesyes
ContainersPlus, Pro and Business (containers)an opt-in profile, on a machine with gVisor
Clusters (Lepis)yes (clusters in Cloud)the router in front of your own Postgres (run it yourself)
AI Gateway (Psyche)yes, switched on per project (AI gateway)the ai profile
Continuous backups, point-in-time restore, every write in another zone on Proyes (durability)your own (pg_dump, volume backups)
Sleeps when idle, wakes on connectyesno
Plans, the dashboard, status pageyesno

Moving between the two is a database dump and the files, since the schemas are the same (move a database).

The source​

PartRepository
The stack (compose file, gateway, setup)snoutdata/snout-stack
Authsnoutdata/snout-auth
Storagesnoutdata/snout-storage
Image resizingsnoutdata/snout-images
Realtimesnoutdata/snout-realtime
Functionssnoutdata/snout-functions
Containerssnoutdata/snout-containers
Pushsnoutdata/snout-push
SnoutTimesnoutdata/snouttime
Lepis (clusters)snoutdata/snout-lepis
Psyche (AI gateway)snoutdata/snout-psyche
JavaScript clientsnoutdata/snout-client (@snoutdata/client on npm)

What changed, part by part, is in the changelog.