Closed aceptra closed 7 years ago
The storage / search/ indexing bits are abstract enough that we have plugged in Couchbase implementations without touching the Spark code directly. We are using Sync Gateway and bucket shadowing to carry FHIR resources to/from a Mobile App and also serve them up via Spark to a web portal. We even share code between mobile and server (Xamarin).
Are you using the dot net client in the xamarin client too? There is a PCL version of it.
We currently have a REST API that we're looking to move all or part to FIHR.
It already uses Couchbase with the .NET client SDK.
No, on Mobile we use Couchbase Lite for seamless synchronization. We use bucket shadowing to sync the official FHIR bucket with the docs coming in/going out of the mobile app. Our web portal uses the official FHIR REST Endpoints provided by Spark.
CBLite on Mobile <---> Sync Gateway <----> Couchbase <---> Spark <---> Web Portal
Relevant Packages on Android (iOS similar)
`
`
@wilsondy, your Spark bits that are talking to Couchbase (not mongodb), are those on GitHub somewhere?
Good to read that MongoDB is sufficiently abstracted away. Still we'll be making the line between MongoDB specific parts and Spark generic parts cleaner in the near future.
I see some history in comments about proposal to connect Spark to Couchbase (@wilsondy ) or other backend.
What is the current status of this and what would it take to change the storage to Couchbase? We are more familiar with Couchbase @here.
Thank you!