Robin Arnold

Refactor SingleResourceResult to avoid returning resource since it will never be updated by PL any more

1

Persistence methods no longer modify the resource (the resource must be fully formed before it hits the persistence layer. This means that there's no reason for the method to return...

A new LRUCache was implemented for 1366 with specific behavior needed for the feature. This should be unified with the other implementation in fhir.core.util.

Use non-XA datasources unless XA is really required

1

**Is your feature request related to a problem? Please describe.** Use of XA-enabled datasources is baked into the proxy XA datasource mechanism. Issue #916 supports the use of traditional JEE...

The current schema Migration class is based around IDatabaseStatement. This makes it difficult to cleanly handle differences among database types. We have now hit several cases where the migration steps...

**Describe the bug** The SHARDED variant of the schema is designed to use an explicit shard key passed in as an HTTP header and injected into the request and persistence...

**Describe the bug** Before the `$erase` operation was implemented there was no way a logical resource record could be deleted by the server. The create/update locking mechanism in the add_any_resource...

PostgreSQL container image should use schema tool for schema creation

1

**Is your feature request related to a problem? Please describe.** Technical debt. **Describe the solution you'd like** Currently the postgres container image hosting the FHIRDB database uses a build script...

**Is your feature request related to a problem? Please describe.** Currently the LinuxForHealth FHIR Server does not enforce uniqueness for identifier values - multiple resources can therefore have the same...

**Is your feature request related to a problem? Please describe.** fhir-bucket can be used to perform reindexing, and provides a client-driven approach. If the process is stopped before completion, it...

**Is your feature request related to a problem? Please describe.** When the sharding feature is ready, we need to advertise the new header: "X-FHIR-SHARD-KEY" in both https://ibm.github.io/FHIR/Conformance#http-headers and https://ibm.github.io/FHIR/guides/FHIRServerUsersGuide#54-custom-http-headers **Describe...