I have configured Azure API for FHIR server and i am able to push some data into it. Using post man i am able to query the server and read the json files.
Now, i would like to move on to the next step. Moving the data into datalake and then apply some machine learning model on it.
First of all - Accessing the FHIR data in Microsoft Azure Storage explorer using the managed version - As per this comment, the data cannot be viewed in Cosmos DB when managed version is used.
Now, i am really confused on where the FHIR-server data is stored. I have submitted few patient information and i am not really sure whether its stored in DB or not.
I have couple of other questions in mind. To get the data out of FHIR server, should i first export it to cosmos Db and then setup a data factory to copy the data into datalake or is there any way, i can directly create a pipeline kind of stuff where data gets into datalake from fhir-serve itself.
Can someone guide on this ?
This question seems to have considerable overlap with your other question. You would use $export to export the data from the FHIR API to some other form of storage, e.g. BLOB storage and then orchestrate whatever pipeline makes sense from there.
Behind the managed FHIR API there is a database, of course, but there is no direct access to that data store in the managed service. If you want to look at the database schema, I would suggest having a look at the open source version of the service (https://github.com/Microsoft/fhir-server). However, I would recommend against taking any specific dependency on the schema as it may change.