Hi folks! I’ve been running Geoserver in Kubernetes with some success, but I’ve been hoping to simplify things. Do you run it just on an instance, or in some other way? I presume a cold start would be rough, so I think I’m avoiding something like Cloud Run or Fargate.
Love to hear what works for everyone!
I think most people are running Cloud-Native geoserver in K8s.
Mm, I should look at that again. Thanks.
One issue I have run into in my current setup is when the monolith Geoserver pod gets moved to another node, the data dir PV is often stuck attached to the previous node and can’t be attached to the new pod. Maybe GeoServer Cloud handles the data dir stuff in a different way?
If I understand correctly, I might want to migrate from a data dir to using pgconfig. or maybe I just need to pin the workload (more of a Kubernetes issue, just looking to see if it’s a common problem for folks running GeoServer on k8s.)
Hi,
I suggest you to go for GeoServer Cloud for your Kubernetes Deployment, with a in-database configuration. If you need an extension which is not currently available in GS Cloud, don’t hesitate to create an issue on the GitHub repo or to make a PR with the wished extension integration!
Regards
Alexandre
Thanks for the advice! I don’t see a clear way to migrate from an existing datadir to an in-database configuration. Is that something I’d have to do myself, or is there a tool for it?