I use a Meilisearch Docker image as a flycast for a webapp.
It uses an internal DB as persistent storage. Storage — Meilisearch documentation
I’m reading through the docs and realized it’s a problem to work with volumes through multiple machines and regions.
I’ve successfully deployed one machine with the volume but I’m wondering how to scale to more machines redundancy and anothers regions. From the docs looks like there’s no solution since the app won’t replicate the data.
Is there any strategy could I try?
It’s not only a question about Meilisearch, it’s a general question about everyone that wants to deploy a docker image that uses persistent volumes and want to scale the server.
Sorry. We don’t have a database-agnostic solution right now.
For SQLite, there is LiteFS - Distributed SQLite · Fly Docs. For Postgres, we have Fly Postgres. But Meilisearch uses its own database engine which we don’t support.
Kaz, my question is more about not being able to scale docker images apps that uses volume. For example, need to deploy an app that needs to mount images to /app/upload/images.
So, if I scale to two machines the second one won’t be synced to the first one. So no solution for scale here.