User Guide

Environments

Below is an image of a single environment. The following will describe what the yaml, lockfile, archive, and docker represent.

Pinned YAML

A pinned YAML file is generated for each environment is built. This includes pinning of the pip packages as well. Note that there are cases where the completely pinned packages do not solve. Packages are routinely marked as broken and removed. Note however conda-forge has a policy that packages are never removed but are marked as broken. Most channels do not obey this policy. When you click the yaml button a YAML file will then be downloaded. To install the environment locally run the following.

conda env create -f <environment-filename>

Conda lockfile

A Conda lockfile is a representation of only the Conda dependencies in a given environment. The lockfile feature was inspired from conda-lock. This file will not reproduce the pip dependencies in a given environment. When working with Conda it is generally not a good idea to mix Conda and pip dependencies. Click the lockfile icon to download the lockfile. First install conda-lock if it is not already installed.

conda install -c conda-forge lockfile

Install the locked environment file from conda-store.

conda-lock install <lockfile-filename>

Conda-Pack archive

Conda-Pack is a package for creating tarballs of given Conda environments. Creating a Conda archive is not as simple as packing and unpacking a given directory. This is due to the base path for the environment that may change. Conda-Pack handles all of these issues. Click the archive button and download the given environment. The size of the archive will be less than the size seen on the environment UI element due to compression.

conda install -c conda-forge conda-pack

Install the Conda-Pack tarball. The directions are slightly complex. Note that my_env can be any name in any given prefix.

mkdir -p my_env
tar -xzf <conda-pack-tarfile>.tar.gz -C my_env

source my_env/bin/activate

conda-unpack

Docker Registry

conda-store acts as a docker registry which allows for interesting ways to handle Conda environment. In addition this registry leverages conda-docker which builds docker images without docker allowing for advanced caching, reduced image sizes, and does not require elevated privileges. Click on the docker link this will copy a url to your clipboard. Note the beginning of the url for example localhost:5000/. This is required to tell docker where the docker registry is located. Otherwise by default it will try and user docker hub. Your url will likely be different.

The conda-store docker registry requires authentication via any username with password set to a token that is generated by visiting the user page to generate a token. Alternatively in the conda_store_config.py you can set c.AuthenticationBackend.predefined_tokens which have environment read permissions on the given docker images needed for pulling.

docker login -u token -p <conda-store-token>
docker pull <docker-url>
docker run -it <docker-url> python

General usage

docker run -it localhost:5000/<namespace>/<environment-name>

If you want to use a specific build (say one that was built in the past and is not the current environment) you can visit the specific build that you want in the UI and copy its docker registry tag name. The tag name is a combination of <specification-sha256>-<build date>-<build id>-<environment name> that we will refer to as build key. An example would be localhost:5000/filesystem/python-numpy-env:583dd55140491c6b4cfa46e36c203e10280fe7e180190aa28c13f6fc35702f8f-20210825-180211-244815-3-python-numpy-env.

docker run -it localhost:5000/<namespace>/<environment-name>:<build_key>

On Demand Docker Image

conda-store has an additional feature which allow for specifying the packages within the docker image name itself without requiring an actual environment to be created on the conda-store UI side.

The following convention is used <registry-url>:<registry-port>/conda-store-dynamic/. After conda-store-dynamic you specify packages needed separated by slashes. Additionally you may specify package constraints for example <=1.10 as .lt.1.10.

As full example support we want python less than 3.8 and NumPy greater than 1.0. This would be the following docker image name. <registry-url>:<registry-port>/conda-store-dynamic/python.lt.3.8/numpy.gt.1.0. conda-store will then create the following environment and the docker image will download upon the docker image being built.

conda-store UI

/ Home Page

The home page shows all of the available environments in the form <namespace>/<environment-name>. If you are authenticated there with be a User button in the top right hand corner to view information about the currently logged in user. Otherwise there is a login button and few if any environments will be visible. Additionally there is a convenient Create Environment button to easily create a given environment. There is a Docs button that will take you to this documentation at any time.

Shortcuts are available below each of the available environments that allow you to download or view the Conda lockfile, YAML, or Conda-Pack files.

/login/ Login

If you are unauthenticated there is a login button on the top navigation bar. This will direct you to the login page. The example above shows what you will get with JupyterHub authentication.

/user/ User

Once a user has completed the authentication flow they will be directed to the user page. This page gives information about the current authenticated user along with the permissions.

/create/ Create Environment

A user authenticated or unauthenticated has set permissions that allow the user to create environments in a given namespace. Currently the create page allows for dragging and uploading of current environment.yaml files in the UI. The specification is the format of a traditional environment.yaml and will report errors if there are issues with the format of the environment file.

/environment/<namespace>/<name>/ Environments

The environment page contains a lot of information for the developer. First we see the environment name and namespace along with the disk space that the environment consumes. Right below this information is the full environment.yaml specification of the currently active build. Users can quickly edit this existing environment by clicking the edit button. Additionally the entire environment can be deleted via the delete button.

Below this is a list of the current builds of the given environment. The environment highlighted in green is the current build for the given environment that is active. The environment highlighted in grey indicates that the build was deleted. Even though a build may be deleted the logs, lockfile, and a other build information is preserved for the record.

For each build several options are available to the user:

• The checkmark icon allow the user to switch that given build to the active build for the environment. This may be useful if you need to rollback a given environment if the new build environment caused some scripts to fail.

• The refresh icon indicates that a user would like the given environment to build again. Conda environment.yaml files are not reproducible thus this will likely lead to an entirely new solve. This is useful when you would like to update all the packages in a given environment without having to change the specification.

• The trashcan icon marks the given build for deletion. CondaStore.build_artifacts_kept_on_deletion allows some artifacts to be kept on deletion. These include logs, YAML, etc.

/build/<build_id> builds

The build page gives all the information about a given build in conda-store. At the top we see high level build metadata.

conda-store downloads Conda channel data so that it fully understands the packages that exist within a given environment. A list is provided to the user of all packages within that environment.

Below this are all artifacts associated with a given build e.g lockfile, pinned YAML specification, Conda-Pack, and docker image.

Finally a log of the given build regardless of whether the build succeeded or failed.

/namespace/ manage namespaces

This namespace page allows a user with correct permissions to list, create, and delete namespaces. Note that the deletion of a namespace is destructive and deletes all environments and builds within that namespace.

conda-store cli

The conda-store client can be easily installed via pip and conda.

pip install conda-store
conda install -c conda-forge conda-store

The base cli is inspired by tools such as conda, kubectl, and docker. The base commands are download, info, list, run, wait.

$ conda-store --help
Usage: conda-store [OPTIONS] COMMAND [ARGS]...

Options:
  --conda-store-url TEXT     conda-store base url including prefix
  --auth [none|token|basic]  conda-store authentication to use
  --no-verify-ssl            Disable tls verification on API requests
  --help                     Show this message and exit.

Commands:
  download  Download artifacts for given build
  info      Get current permissions and default namespace
  list
  run       Execute given environment specified as a URI with COMMAND
  solve     Remotely solve given environment.yaml
  wait      Wait for given URI to complete or fail building

conda-store run

One of the motivating features of the conda-store cli is that you can directly execute conda-store environments that exist remotely.

conda-store run devops/datascience -- python -m "print(1)"

conda-store solve

conda-store is capable to remote solves of environment files. If requested conda-store can perform intelligent solves with caching.

conda-store download

conda-store info

conda-store wait

conda-store list [namespace|environment|build]

conda-store shebang

conda-store can be used as a shebang within Linux allowing users to embed Conda environments within scripts for reproducibility. Basic usage is as follows. Notice that the conda-store run command is just the normal usage of the command.

#!/usr/bin/env conda-store
#! conda-store run <namespace>/<environment-name>:<build-id> -- python

print('script running within the conda-store environnent')

The first line must begin with the shebang #! along with ending in conda-store. You cannot put arguments on the first line due to limits in the shebang specification. Additional lines are then added starting with #! conda-store run ... with are then used as arguments to conda-store run.

The path to the script being run is always appended as the last argument to the command so the example above is interpreted as:

conda-store run <namespace>/<environment-name>:<build-id> -- python <shebang-filename>

This feature was heavily inspired by nix-shell shebangs.