Frequently Asked Questions (FAQ)#

Welcome to the FAQ section of the freva admin documentation. Here you’ll find answers to common questions, troubleshooting tips, and practical guidance related to setting up, configuring, and running the Freva framework.

Whether you’re deploying services with Docker, Podman, or Conda environments, this page is here to help resolve typical issues and clarify common concerns encountered during the installation and operation phases.

This section is actively maintained and expands over time as new questions arise from users and developers. If your question isn’t covered yet, feel free to open an issue or contribute a new entry.

Topics covered#

  • Service initialization problems

  • Database and storage configuration

  • Container runtime compatibility

  • Secrets and environment variable handling

  • Version pinning and updates

  • Integration with monitoring and orchestration tools


🐳 Why is docker compose failing with “command not found”?#

If you’re using Docker but docker compose fails, it’s likely that the Docker Compose plugin isn’t installed. Modern Docker uses docker compose (with a space), not docker-compose (hyphen).

  • On Debian/Ubuntu:

    sudo apt install docker-compose-plugin
    
  • On RHEL/AlmaLinux

    sudo dnf install docker-compose-plugin
    

If you’re using Podman, ensure podman-compose is installed and available in your $PATH.

📦 Where are logs and volumes stored?#

Depending whether you’ve chosen conda-forge based or a docker/podman based deployment approach your logs data data will be located in diffrent locations:

  • conda-forge: All data will be stored in <data_dir>/<project_name>/services/<service>

  • podman/conda: The data will be located in container volumes managed by the container engine checkout

    docker volume ls
    docker volume inspect <project_name>-<service>_data
    

    Or using podman if the service was deployed with podman.

💥 Can’t inject secrets into vault.#

If secret injection fails but the <project_name>-vault is up and running you can check the logs. In most cases there will be a verion mismatch of the a older deployed version and the current vault version. To overcome this stop the service and delte the data:

For conda-forge based deployments:

rm -r <data_dir>/<project_name>/services/vault

Or for podman/docker:

docker volume rm -f <project_name>-vault_data