Performance studio backend and frontend

February 29, 2024 ยท View on GitHub

Running the backend locally

Pre-requisites

  • Python 3.12 or higher is installed on your machine.
  • Running deploy/docker-compose.yml without the webapp service - that way you will be covered with all the dependencies, like databases and other services.

Now we need to install the dependencies:

cd src/gprofiler
pip install -e ../gprofiler-dev[postgres]
pip install -e .

Running the backend

To run the backend, you need to set the following environment variables:

  • BUCKET_NAME - S3 bucket name
  • QUERY_API_BASE_URL - running src/gprofiler_flamedb_rest http://localhost:8080
  • SQS_INDEXER_QUEUE_URL - SQS queue name
  • GPROFILER_POSTGRES_DB_NAME
  • GPROFILER_POSTGRES_PORT
  • GPROFILER_POSTGRES_HOST
  • GPROFILER_POSTGRES_USERNAME
  • GPROFILER_POSTGRES_PASSWORD
  • APP_LOG_FILE_PATH - path to the application log file
  • APP_LOG_LEVEL
  • AWS_METADATA_SERVICE_NUM_ATTEMPTS - number of attempts to get the AWS credentials
  • REDIRECT_DOMAIN - domain name for the installation instruction generation in UI
  • SQS_ENDPOINT_URL - SQS endpoint URL, by default https://sqs.${AWS_REGION}.amazonaws.com
  • AWS_ACCESS_KEY_ID - AWS credentials, if empty is taken from the metadata service
  • AWS_SECRET_ACCESS_KEY
  • AWS_SESSION_TOKEN

Now you can run the backend:

python3 -m uvicorn backend.main:app --reload --port 8000

Running the frontend locally

cd src/gprofiler/frontend
yarn install
yarn start