More from the engine
Atlas edits and checks the mappings; the openFHIR engine runs them. The engine can do more than Atlas shows on screen. This page lists those capabilities, with a link to where the engine documentation describes each one.
You want to… |
The engine offers |
Read more |
|---|---|---|
Convert data that already travels over a message broker |
Queue support (Kafka): the engine subscribes to topics carrying FHIR payloads or openEHR compositions, translates every message and publishes the result to another topic. No REST-calling consumer needed in between. |
|
Run several engine nodes behind a load balancer |
High availability: a change made through one node (a new mapping or OPT) reaches the caches of all the others automatically, without extra infrastructure. |
|
Query an openEHR CDR with FHIR searches |
AQL generation: the engine turns a FHIR request such as |
|
Run logic a path can’t express |
Programmed mappings: a mapping’s Mapping code names a function built into the engine. The engine ships narrative generation ( |
|
Call the engine from your systems |
REST and FHIR operations: |
|
Load mappings without calling the API |
Bootstrap: OPTs, context and model mappers (and ConceptMaps) placed in a directory are loaded when the engine starts, and again on |
Tip
Atlas covers the authoring side of all of these: the mappings you write here are what the queue consumer, AQL generation and the REST operations run. API access creates the client your system uses for the REST calls (Connect your system to the API).
Other engine topics that matter when you plan an installation: Multitenancy, Integrations (openEHR CDR, terminology server), Caching, Authentication and Mapping insights.