tw
is Seqera Platform on the command line. It brings Seqera concepts like pipelines, actions, and compute environments to the terminal.
Seqera Platform is a full-stack application for the management of data pipelines and compute resources. It enables collaborative data analysis at scale, on-premises or in any cloud.
The CLI interacts with Seqera to provide an interface to launch pipelines, manage cloud resources, and administer your analysis.
The key features are:
-
A Nextflow-like experience: tw CLI provides a developer-friendly environment. Pipelines can be launched with the CLI similarly to Nextflow but with the Seqera benefits of monitoring, logging, resource provisioning, dataset management, and collaborative sharing.
-
Infrastructure as Code: All Seqera resources, including pipelines and compute environments, can be described in a declarative manner. This enables a complete definition of an analysis environment that can be versioned and treated as code. It greatly simplifies configuration sharing and routine administration.
-
Built on OpenAPI: tw CLI interacts with Seqera via the Seqera Platform API which uses the OpenAPI 3.0 specification. The CLI provides full control of the Seqera application, allowing users to get maximum insights into pipeline submissions and execution environments.
See the Seqera Platform documentation to learn more about the application.
tw CLI can be installed on macOS, Windows, and Linux.
It is compatible with Seqera Cloud and Enterprise versions 21.08 and later.
This guide covers the installation and configuration of the CLI, cloud infrastructure provisioning, and launching pipelines into an AWS Batch compute environment.
-
Download the latest version for your OS from the assets on the releases page.
-
Rename and make the file executable:
mv tw-* tw chmod +x ./tw
-
Move the file to a directory accessible to your
$PATH
variable:sudo mv tw /usr/local/bin/
You need an access token for the CLI to interact with your Seqera instance. Select User tokens from the user menu in the Seqera UI, then select Add token to create a new token.
Copy the access token value and use it with the CLI in one of two ways:
-
Environment variable:
-
Export the token as a shell variable directly into your terminal:
export TOWER_ACCESS_TOKEN=<your access token>
-
Add the
export
command to your.bashrc
,.zshrc
, or.bash_profile
file for it to be permanently added to your environment.
-
-
tw command flag:
Provide the access token directly in your
tw
command with--access-token
:tw --access-token=<your access token> <other options>
If required, configure the following non-mandatory environment variables using the same methods above:
TOWER_WORKSPACE_ID
: Workspace ID. Default: The user workspace.TOWER_API_ENDPOINT
: Seqera API URL. Default:api.cloud.seqera.io
.
You can find your
TOWER_WORKSPACE_ID
from the Workspaces tab on your organization page. Alternatively, you can list all the workspaces your token can access withtw workspaces list
and copy the workspace ID from the command output.
Confirm the installation, configuration, and connection:
$ tw info
System health status
---------------------------------------+------------------
Remote API server connection check | OK
Tower API version check | OK
Authentication API credential's token | OK
See Usage for detailed instructions to use the CLI.
Activate autocompletion in your current session with this command:
source <(tw generate-completion)
If you are using a Private CA SSL certificate not recognized by the default Java certificate authorities, use a custom cacerts
store:
tw -Djavax.net.ssl.trustStore=/absolute/path/to/cacerts info
You can rename the binary to tw-binary
and create a tw
script to automatically include the custom cacerts
store in every session:
#!/usr/bin/env bash
tw-binary -Djavax.net.ssl.trustStore=/absolute/path/to/cacerts $@
tw CLI is a platform binary executable created by a native compilation from Java GraalVM. To compile and build a development version of the binary:
-
If necessary, install SDKMan!
-
From the root of the tower-cli project, install GraalVM:
sdk env install
This ensures that SDKMan uses the tower-cli project-specific
.sdkmanrc
configuration. -
Install
native-image
:gu install native-image
-
Export your Github credentials. Github requires authentication for public packages (the token only requires the
read:packages
scope):export GITHUB_USERNAME=... export GITHUB_TOKEN=...
-
Create the native client:
./gradlew nativeCompile
This will install a locally compiled version of
tw
in the nativeCompile directory:Produced artifacts: <tower-cli-repository-root>/build/native/nativeCompile/tw (executable) ======================================================================================================================== Finished generating 'tw' in 1m 6s. [native-image-plugin] Native Image written to: <tower-cli-repository-root>/build/native/nativeCompile BUILD SUCCESSFUL in 1m 8s 6 actionable tasks: 2 executed, 4 up-to-date
-
Run
tw
:./build/native/nativeCompile/tw
Run a non-binary development version by executing the ./tw
script in the root of this repository.