Configuration
Torch is configured using environment variables that control its behavior and integration with a FHIR server. This allows for flexible deployment and customization based on the specific requirements of your environment.
Configuration Overview
This document provides an overview of the environment variables used to configure TORCH, including their default values and descriptions. These variables can be set in the application.yml file during development or as system environment variables when running the application.
When running TORCH as a container, it is recommended to set these variables or pass them as environment variables during container startup.
Backend
Environment Variables
SERVER_PORT Since 1.0.0-alpha
The port on which TORCH server listens for incoming connections. If running a container it is the port inside the container.
Default: 8080
TORCH_BASE_URL Since 1.0.0-alpha
Base server URL before any proxy, used for accessing TORCH directly.
Default: – http://localhost:8080
TORCH_PROFILE_DIR Since 1.0.0-alpha
The directory path where profile definitions are stored.
Default: structureDefinitions
TORCH_MAPPING_CONSENT Since 1.0.0-alpha
Path to the JSON file containing consent mappings formatted for FHIR.
Default: mappings/consent-mappings_fhir.json
TORCH_MAPPING_TYPE_TO_CONSENT Since 1.0.0-alpha
File that maps FHIR Resource Types to their associated time fields used to check consent validity.
Default: mappings/type_to_consent.json
TORCH_FHIR_USER Since 1.0.0-alpha
The username used for authentication with the FHIR server.
Default: – (none)
TORCH_FHIR_PASSWORD Since 1.0.0-alpha
The password used for authentication with the FHIR server.
Default: – (none)
TORCH_FHIR_OAUTH_ISSUER_URI Since 1.0.0-alpha
The URI of the OAuth issuer used for OAuth authentication.
Default: – (none)
TORCH_FHIR_OAUTH_CLIENT_ID Since 1.0.0-alpha
Client ID used for OAuth authentication.
Default: – (none)
TORCH_FHIR_OAUTH_CLIENT_SECRET Since 1.0.0-alpha
Client secret used for OAuth authentication.
Default: – (none)
TORCH_FHIR_URL Since 1.0.0-alpha
Base URL of the FHIR server that TORCH connects to.
Default: – (none)
TORCH_FHIR_MAX_CONNECTIONS Since 1.0.0-alpha
Maximum number of concurrent connections allowed to the FHIR server. Must be at least TORCH_MAXCONCURRENCY + 1.
Default: 5
TORCH_FHIR_PAGE_COUNT Since 1.0.0-alpha
Number of entries per page in FHIR search responses.
Default: 500
TORCH_FHIR_DISABLE_ASYNC Since 1.0.0-alpha
Set to true to disable the use of the Asynchronous Interaction Request Pattern for FHIR operations.
Default: false
TORCH_FLARE_URL Since 1.0.0-alpha
Base URL of the FLARE server used in the pipeline.
Default: – (none)
TORCH_RESULTS_DIR Since 1.0.0-alpha
Directory path where results are stored.
Default: output/
TORCH_RESULTS_PERSISTENCE Since 1.0.0-alpha
ISO 8601 duration indicating how long result files are persisted (e.g., PT2160H means 90 days).
Default: PT2160H
TORCH_BATCHSIZE Since 1.0.0-alpha
Size of data batches processed at once.
Default: 500
TORCH_MAXCONCURRENCY Since 1.0.0-alpha
Maximum level of concurrency for data processing operations.
Default: 4
TORCH_MAPPINGS_FILE Since 1.0.0-alpha
Path to the file containing ontology mappings defined using Clinical Quality Language (CQL).
Default: ontology/mapping_cql.json
TORCH_BUFFERSIZE Since 1.0.0-alpha
Size (in MB) of the buffer used by the web client interacting with the FHIR server.
Default: 100
TORCH_CONCEPT_TREE_FILE Since 1.0.0-alpha
File containing the concept tree mapping used for resource classification.
Default: ontology/mapping_tree.json
TORCH_DSE_MAPPING_TREE_FILE Since 1.0.0-alpha
File containing the concept tree mapping specifically for DSE (Data Set Extensions).
Default: ontology/dse_mapping_tree.json
TORCH_USE_CQL Since 1.0.0-alpha
Flag indicating whether to enable Clinical Quality Language (CQL) for processing.
Default: true
TORCH_OUTPUT_FILE_SERVER_URL Since 1.0.0-alpha
URL used to access the result files in TORCH_RESULTS_DIR via a proxy or file server.
Default: – (none)
LOG_LEVEL_DE_MEDIZININFORMATIKINITIATIVE_TORCH Since 1.0.0-alpha
Logging level for core TORCH functionality (error/ warn/ info/ debug/ trace).
Default: info
LOG_LEVEL_CA_UHN_FHIR Since 1.0.0-alpha
Logging level for the HAPI FHIR library (error/ warn/ info/ debug/ trace, fatal, off).
Default: error
SPRING_PROFILES_ACTIVE Since 1.0.0-alpha
Active Spring profile used to configure the application context (only needed for development).
Default: active
SPRING_CODEC_MAX_IN_MEMORY_SIZE Since 1.0.0-alpha
Maximum allowed in-memory size for Spring codecs when processing data.
Default: 100MB
Monitoring Since 1.0.0-alpha.13
Torch allows monitoring using the Spring boot default actuator endpoints.
Custom metrics might be added with time.
For an example setup see application-test.ymlNote: Be Careful with exposing these endpoints outside