OMERO Server - docker compose

I’ve been using the OMERO server docker-compose file for testing updates to Orbit Image Analysis, and it has really saved me a lot of time so far. Thanks!

I also built a simple test running on AWS using Fargate to manage the containers. I was wondering if anyone has any experience of running OMERO in a ‘productive’ environment?

We have lots of examples of running OMERO in production on virtual machines, including the Image Data Resource and the University of Dundee’s public OMERO.web https://omero.lifesci.dundee.ac.uk/. You should be able to find more examples in this forum, including this thread:
Omero use world wide?

I know some people are running in production on Docker but I’m afraid I can’t point to particular examples. Is there anything in particular you’d like to know?

Thanks for the feedback. I’m mainly wondering about considerations of state. E.g. if I want to upgrade from version 5.5 to 5.6, am I likely to be able to simply deregister the 5.5 container and launch a 5.6 container in its place (just chose those two versions as an example, wondering generally)?

Is there state held in the ‘server’ or ‘web’ containers that would prevent running a simple failover setup (for high availability)?

I’m happy to read through some documentation here, but perhaps you could help by directing my reading a little.

All the state is held in PostgreSQL and the OMERO data directory. The omero-server Docker image will automatically upgrade your database schema if required, though there were no schema changes in 5.6 so in your case it won’t apply.

You can have multiple OMERO.web clients connecting to one OMERO.server, but only one OMERO.server can have read-write access to the PostgreSQL database due to a limitation in OMERO. This will prevent you from having high availability with OMERO- you can only start your failover server once the original instance has disconnected from the DB.

Thanks for the great overview, that’s exactly what I needed.