Cron Workflows

Using the Pipekit CLI with CronWorkflows

The Pipekit CLI provides create (equivalent to submit for regular workflows), delete, suspend, resume, and update commands for a CronWorkflow.

For further information about CronWorkflows, see our CronWorkflow documentation.

Create and update Cron

To create a new CronWorkflow on a given cluster:

$> pipekit create cron --cluster-name=clusterB ./path/to/workflow.yaml

This will create a new CronWorkflow and a Pipe based on the CronWorkflow name (part of the metadata inside the YAML). If such a Pipe already exists, a new CronWorkflow manifest will be associated with it as well.

Note: You will not be able to create two CronWorkflows with the same name, on the same cluster, in the same namespace. You must therefore check the namespace specified in the meta part of your CronWorkflow yaml.

To update an existing CronWorkflow, on a given cluster:

$> pipekit update cron --cluster-name=clusterB ./path/to/workflow.yaml

Flag	Shorthand	Type	Description
--cluster-name	-c	string	name of the cluster the CronWorkflow lives on (required)
--namespace	-n	string	namespace in which the CronWorkflow is located

Delete Cron

To delete a CronWorkflow, from a given cluster, in a given namespace:

$> pipekit delete cron --cluster-name=clusterB --namespace=foo cron-wf-name

The argument passed to this command is the name of the CronWorkflow (which should correspond to the Pipe name).

Flag	Shorthand	Type	Description
--cluster-name	-c	string	name of the cluster from which to delete (required)
--namespace	-n	string	namespace in which the CronWorkflow is located

Suspend and resume Cron

Suspending a CronWorkflow means to stop it from scheduling any new Workflows. The Pipekit CLI lets you suspend a specific CronWorkflow in a given cluster and namespace, without suspending all CronWorkflows in a given Pipe.

See Disabling a Pipe for suspending at the Pipe level in the UI.

To suspend a CronWorkflow, on a given cluster, in a given namespace:

To resume a suspended CronWorkflow, on a given cluster, in a given namespace:

The only argument that these commands accept is the name of the CronWorkflow to suspend/resume.

Flag	Shorthand	Type	Description
--cluster-name	-c	string	name of the cluster from which to suspend (required)
--namespace	-n	string	namespace in which the CronWorkflow is located

Note: The only way to suspend and resume a specific CronWorkflow, on a given cluster, in a given namespace without suspending or resuming the CronWorkflows on other clusters and namespaces is to use the Pipekit CLI. The toggle for a Pipe in the Pipekit web interface will suspend and resume ALL CronWorkflows within the Pipe.

Trigger Cron

To run a CronWorkflow immediately, on a given cluster, in a given namespace:

The only argument that this command accepts is the name of the CronWorkflow to trigger.

Flag	Shorthand	Type	Description
--cluster-name	-c	string	name of the cluster on which to trigger the run (required)
--namespace	-n	string	namespace in which the CronWorkflow is located
--open-ui		bool	open the Pipekit web interface in your browser for the submitted workflow run
--wait	-w	bool	wait for the submitted workflow to complete