Conda Repositories

JFrog Artifactory natively supports Conda repositories for Python, R, Ruby, Lua, Scala, Java, JavaScript, C/C++, FORTRAN, and additional programming languages, giving you full control of your deployment and resolution process of Conda packages. Artifactory supports Conda client version 4.12.0 and above, and using the latest version of the client is recommended.

Conda repositories streamline package management by offering secure local hosting with fine-grained access control, proxying and caching remote packages for network independence, and aggregating all resources under a single virtual URL with built-in metadata calculation and version management.

For more information, see the Conda documentation.

Get Started with Conda

To get started working with Conda, complete the following main steps:

1. Create a Conda Repository
2. Connect Conda to Artifactory
3. Deploy and Resolve Conda Packages

Create a Conda Repository

Create a Conda repository before you publish or install Conda packages. The primary repository types are:

• Local repositories: Where you store and share 1st and 2nd party packages with your organization
• Remote repositories: Enable you to download from any remote location, including external package registries or other Artifactory instances
• Virtual repositories: Enable aggregating remote and local repositories, enabling your organization to scale by providing a single URL that provides access to multiple repositories and types

For more information, see Repository Management.

Prerequisite: You need Admin or Project Admin permissions to create a Conda repository. If you don't have Admin permissions, the option will not be available.

To create a Conda repository:

1. In the Administration module, click Repositories > Create a Repository.

   

2. Select the repository type you want to create.

3. Select the Conda package type.

4. Configure the required fields:

   • For Local Repositories, in the Repository Key field, type a meaningful name for the repository. For example, conda-local-repository-1. For more information, see Local Repositories.

   • For Remote Repositories, verify the Repository URL and update if needed. For more information, see Remote Repositories.

• For Virtual Repositories, select one or more local or remote repositories to include in the virtual repository. For more information, see Virtual Repositories.

5. Click Create Repository. The repository is created, and the Repositories page is displayed.

Connect Conda to Artifactory

To resolve Conda packages from Artifactory, configure your .condarc file to connect the Conda client to Artifactory.

Prerequisite: Before connecting your Conda client to Artifactory, you must have an existing Conda repository in Artifactory. For more information, see Create a Conda Repository.

To connect the Conda client to Artifactory:

1. Open your .condarc file in a text editor. If the file does not exist, create it in your home directory.

2. Add the following snippet to configure Artifactory as your Conda repository:

   YAML

   channel_alias: https://<URL_ENCODED_USERNAME>:<AUTH>@<JFrogPlatformURL>/artifactory/api/conda/<REPO_NAME>
   channels:
     - https://<URL_ENCODED_USERNAME>:<AUTH>@<JFrogPlatformURL>/artifactory/api/conda/<REPO_NAME>
   default_channels:
     - https://<URL_ENCODED_USERNAME>:<AUTH>@<JFrogPlatformURL>/artifactory/api/conda/<REPO_NAME>

   Where:

   • <URL_ENCODED_USERNAME>: Your Artifactory username, with @ encoded as %40 if present
   • <AUTH>: Your Artifactory identity token
   • <JFrogPlatformURL>: Your JFrog Platform URL
   • <REPO_NAME>: The name of the target repository

   For example:

   YAML

   channel_alias: https://jeffry:[email protected]/artifactory/api/conda/conda-local
   channels:
     - https://jeffry:[email protected]/artifactory/api/conda/conda-local
   default_channels:
     - https://jeffry:[email protected]/artifactory/api/conda/conda-local

3. (Optional) To disable Conda client SSL validation, add the following directive to your .condarc file:

   YAML

   ssl_verify: false

Next steps:

• Deploy Conda Packages
• Resolve Conda Packages

Deploy Conda Packages

Deploying Conda packages makes them available to other users.

The following options are available for publishing Conda packages:

• Deploy Conda Package Using the UI
• Deploy a Conda Package Using cURL

Deploy Conda Packages Using the UI

Upload Conda packages through the Artifactory web UI when you prefer a graphical workflow.

To deploy a Conda package using the UI:

1. In the Platform module, click Artifactory > Artifacts. The Artifacts page opens.

2. Click the actions button and select Deploy (Upload). Deploy options are displayed.

   

3. Select your Conda repository as the Target Repository, and upload the file you want to deploy.

4. Select the Deploy as Conda Artifact checkbox. When deploying, ensure you upload the package to the correct Target Path that reflects its subdirectory, for example, your-repository/linux-64/package-name.tar.bz2. Artifactory uses this path to correctly index the package.

5. Verify the deploy options, then click Deploy. The package is deployed.

Deploy Conda Packages Using cURL

You can deploy a Conda package using cURL by uploading the package file to the correct path in your repository.

To deploy a Conda package using cURL:

• Run the following command to upload the package to your Artifactory repository:

  Shell

  curl -L -u <USERNAME>:<PASSWORD> -T <PATH_TO_FILE> "https://<JFrogPlatformURL>/artifactory/<REPO_NAME>/<TARGET_FILE_PATH>"

  Where:

  • <USERNAME>: Your Artifactory username
  • <PASSWORD>: Your Artifactory password or identity token
  • <PATH_TO_FILE>: Local path to your package file
  • <JFrogPlatformURL>: Your JFrog Platform URL
  • <REPO_NAME>: The name of the target repository
  • <TARGET_FILE_PATH>: Remote path in the repository. Use <SUBDIR>/<PACKAGE_FILE_NAME> (for example, linux-64/sample-1.0.0-py38_0.tar.bz2). To deploy to a Conda sub-channel, prefix the path with the channel name (for example, main/linux-64/sample-1.0.0-py38_0.tar.bz2).

  For example, to deploy a package to the main sub-channel:

  Shell

  curl -L -ujeffry:RandomToken!_51HnvD0A1g2q3n4b5F6g -T ./linux-64/sample-1.0.0-py38_0.tar.bz2 \
  "https://company.jfrog.io/artifactory/conda-local/main/linux-64/sample-1.0.0-py38_0.tar.bz2"

Resolve Conda Packages

You can resolve Conda packages from local, remote, and virtual Conda repositories after you connect the Conda client to Artifactory.

To resolve Conda packages from Artifactory:

1. Run the following command to install a package from Artifactory:

   Shell

   conda install <PACKAGE_NAME>

   Where:

   • <PACKAGE_NAME>: The name of the Conda package to install

   For example:

   Shell

   conda install numpy

2. To install a package from a specific Conda sub-channel, run the following command with the -c flag:

   Shell

   conda install -c <CHANNEL_NAME> <PACKAGE_NAME>

   Where:

   • <CHANNEL_NAME>: The Conda sub-channel URL or name
   • <PACKAGE_NAME>: The name of the Conda package to install

   For example:

   Shell

   conda install -c https://company.jfrog.io/artifactory/api/conda/conda-virtual pandas

Additional Conda Actions

The following additional actions are available with Conda repositories:

• Reindex a Conda Repository
• Tune Conda Metadata Worker Threads
• View Individual Conda Package Information

Reindex a Conda Repository

You can trigger asynchronous reindexing of a local Conda repository either through the UI or using the REST API.

To reindex a Conda repository:

1. In the Platform module, go to Artifactory > Artifacts.
2. Find your repository in the repository tree, right-click it, and select Recalculate Index.

For more information, see Calculate Conda Repository Metadata API.

Tune Conda Metadata Worker Threads

Conda calculates metadata asynchronously based on repository storage events. The number of total worker threads that handle metadata calculation in parallel (specifically for Conda tasks) defaults to 5. At larger scales, you may modify this parameter in the Artifactory system properties file.

To tune Conda metadata worker threads:

• Edit $JFROG_HOME/artifactory/var/etc/artifactory/artifactory.system.properties and add the following parameter:

  Text

  artifactory.conda.metadata.calculation.workers=<NUMBER_OF_WORKERS>

  Where:

  • <NUMBER_OF_WORKERS>: The number of parallel worker threads for Conda metadata calculation

  For example:

  Text

  artifactory.conda.metadata.calculation.workers=10

View Individual Conda Package Information

Artifactory supports viewing selected Conda package metadata directly from the UI.

To view individual Conda package information:

• In the Platform module, go to Artifactory > Artifacts, expand the repository in the repository tree, and select the package file you want to inspect.

Related Topics

• Connect Your Package Manager to Artifactory
• Repository Management
• Use Artifactory Set Me Up for Configuring Package Manager Clients
• Deploy Artifacts
• Local Repositories