Skip to content
LAPIS

Starting SILO and LAPIS

Starting SILO

SILO can be started in two modes:

  • preprocessing: Starts the SILO preprocessing that prepares the data for the SILO API server.
  • api: Starts the SILO API server that responds to SILO queries.

We provide Docker images for SILO with the name: ghcr.io/genspectrum/lapis-silo. Those images can be used to start SILO in both modes.

Starting the SILO preprocessing

To start the SILO preprocessing you need to provide the argument --preprocessing. SILO will then start reading the input data, process it and terminate after the result is written to the output directory.

Optionally, you can specify:

  • --preprocessingConfig: The path to the preprocessing config file that SILO should use, e.g. ~/LAPIS/preprocessing_config.yaml
  • --databaseConfig: The path to the database config file that SILO should use, e.g. ~/LAPIS/database_config.yaml.

If those parameters are not provided, SILO will try to read the files from its current working directory. See the preprocessing reference for how to provide the input files.

Starting the SILO preprocessing with Docker

You only have to pass the argument --preprocessing to the Docker container. The image is constructed with defaults in a way that you only need to mount the data to the correct locations:

  • the preprocessing config to /app/preprocessing_config.yaml,
  • the database config to /app/database_config.yaml,
  • the data to /preprocessing/input,
  • the output directory to /preprocessing/output.

Starting the SILO API

To start the SILO API server you need to provide the argument --api. It will start a web server that listens to port 8081.

The SILO API will read data from a directory that can be specified (in descending order of prevalence) via:

  • a program argument --dataDirectory,
  • YAML file ./runtime_config.yaml (in SILO’s working directory) with the key dataDirectory,
  • the default value ./output/ (in SILO’s working directory).

The database config does not need to be provided. It is contained in the compiled output of the preprocessing.

Starting the SILO API with Docker

You only have to pass the argument --api to the Docker container. The image is constructed with a runtime config file that sets the data directory to /data, i.e. you only need to mount the result of the preprocessing to /data.

Starting LAPIS

To start LAPIS you need to provide the following parameters:

  • --silo.url: The URL where SILO is running, e.g. http://localhost:8081.
  • --lapis.databaseConfig.path: The path to the database config file that LAPIS should use, e.g. ~/LAPIS/database_config.yaml.
  • --referenceGenomeFilename: The path to the reference genomes file that LAPIS should use, e.g. ~/reference_genomes.json. This can alternatively be passed via the environment variable REFERENCE_GENOME_ENV_VARIABLE_NAME. The parameter takes precedence over the environment variable.

LAPIS listens on port 8080.

LAPIS provides the following optional parameters:

  • --lapis.docs.url: If set, the landing page (/) will show a “Documentation” link to the provided value. It is meant to point to your self-hosted version of this LAPIS documentation. If not set, the “Documentation” link on the landing page (/) will not be shown.

Starting LAPIS with Docker

We provide Docker images for LAPIS with the name: ghcr.io/genspectrum/lapis.

The provided Docker images already contain reasonable default values for

  • the lapis.databaseConfig.path parameter: /workspace/database_config.yaml,
  • the REFERENCE_GENOME_ENV_VARIABLE_NAME environment variable: /workspace/reference_genomes.json.