Command-line interface#

nublado#

Command-line interface for nublado.

Usage

nublado [OPTIONS] COMMAND [ARGS]...

Options

--version#

Show the version and exit.

help#

Show help for any command.

Usage

nublado help [OPTIONS] [TOPIC] [SUBTOPIC]

Arguments

TOPIC#

Optional argument

SUBTOPIC#

Optional argument

inithome#

Provision user home directory.

All of NUBLADO_UID, NUBLADO_GID, and NUBLADO_HOME must be set.

This runs as an initContainer before either the landing page initContainer or the startup initContainer runs.

Usage

nublado inithome [OPTIONS]

landingpage#

Redirect user to writeable copy of CST landing page.

This only occurs at “science”-type RSP instances. It uses the environment variables NUBLADO_HOME, CST_LANDING_PAGE_SRC_DIR, CST_LANDING_PAGE_TGT_DIR, and CST_LANDING_PAGE_FILES. It must never throw an exception: if the init container holding the landing page provisioner fails to perform, we just open a Lab as usual, with whatever layout is in the user cache.

This runs as an initContainer before the startup initContainer runs.

Usage

nublado landingpage [OPTIONS]

purger#

Purge files, or plan future purge.

Usage

nublado purger [OPTIONS] COMMAND [ARGS]...

Options

-x, --dry-run#

Do not act, but report what would be done

-p, --policy-file <policy_file>#

Purger policy configuration file

-c, --config-file <config_file>#

Application configuration file

-d, --debug#

Enable debug logging

Environment variables

DEBUG

Provide a default for --debug

execute#

Make a plan, report, and purge files.

Usage

nublado purger execute [OPTIONS]

Options

-x, --dry-run#

Do not act, but report what would be done

-p, --policy-file <policy_file>#

Purger policy configuration file

-c, --config-file <config_file>#

Application configuration file

-d, --debug#

Enable debug logging

Environment variables

DEBUG

Provide a default for --debug

report#

Report what files would be purged.

Usage

nublado purger report [OPTIONS]

Options

-x, --dry-run#

Do not act, but report what would be done

-p, --policy-file <policy_file>#

Purger policy configuration file

-c, --config-file <config_file>#

Application configuration file

-d, --debug#

Enable debug logging

Environment variables

DEBUG

Provide a default for --debug

warn#

Make a plan for some time in the future, and report as if it were that time.

Usage

nublado purger warn [OPTIONS]

Options

-t, --future-duration <future_duration>#

Duration from now to future time to build a plan for

-x, --dry-run#

Do not act, but report what would be done

-p, --policy-file <policy_file>#

Purger policy configuration file

-c, --config-file <config_file>#

Application configuration file

-d, --debug#

Enable debug logging

Environment variables

DEBUG

Provide a default for --debug

startup#

Prepare environment for launching user Lab pod.

This makes any needed changes to the filesystem (e.g. merging credentials into the user’s homedir), and sets up the runtime environment (e.g. setting the idle-cull timeout); then it dumps those settings somewhere they can be consulted at container start time.

This runs as an initContainer just before the Lab image launches JupyterLab.

Usage

nublado startup [OPTIONS]

Controller#

The Nublado controller doesn’t have a command-line interface. It should be started via uvicorn, using nublado.controller.main:create_app as the entry point, with host 0.0.0.0 and port 8080.

Filesystem Administration#

The filesystem administration function is simply to start the container with root privilege and then keep the container alive while allowing kubectl exec access to the pod. Typically it will be started with an innocuous command such as tail -f /dev/null.

Repository Cloning#

The repository cloner is not started through the nublado command-line interface. Instead, the shell entrypoint will be invoked directly as /usr/local/bin/repo-cloner.sh (typically from a Kubernetes CronJob).