Execute commands

With the nuvolos API you can execute commands in your running Nuvolos applications with the nuvolos apps execute command:

nuvolos apps execute -o <org_slug> -s <space_slug> -i <instance_slug> -a <app_slug> COMMAND

Selecting the application for command execution

You can select the application where you intend to execute the command by providing the app_slug for the -a or --app option, while application context can be provided the -o, -s, -i options.

On Nuvolos, you can omit the -o, -s, -i options if your target application is running in the same instance. However, if you'd like to submit a command to an application that is running in a separate instance, you need to specify the context for that application.

Commands can only be run in a Nuvolos application that is in a RUNNING state. To understand how to verify that your application is running, you can refer to this guide. By default the Nuvolos API runs the command in an application that you started. However, in instances where there is no application initiated by the current user with a matching app_slug, yet a shared application with the same app_slug is running in the same instance — initiated by a different user - the API will proceed to execute the command within that shared application.

Submitting a command for execution

You can submit multiple commands to a running Nuvolos application by making multiple calls the nuvolos apps execute command. The command execution mirrors the behavior of the terminal in your Nuvolos application when accessed through the UI. For example, you can run .py files in a JupyterLab application by calling 

nuvolos apps execute -a your_app_slug 'python your_file.py'

The default working directory for command execution is the same as the interactive terminals in the selected applications, the /files folder.

Storing metadata about the submitted commands

Upon every command execution a new folder is created in selected application's instance under the /files/nuvolos_api_out folder in a <timestamp>_<request_id> format, where the request id (reqid) is returned by the cli command. A metadata.json file is created to store details such as the application context, submission timestamp, and the submitted command.

Accessing standard output and standard error

By default, the standard output and the standard error of your command are redirected to separate files to ensure preservation. The files are located in the folder introduced above, called output.log and error.log, respectively.

Note

nuvolos apps execute supports default output redirect when exactly one command is submitted. If you submitted a command sequence (e.g. python prepare.py && python evaluate.py), you need to specify the files where you intend to redirect the stdout and stderr of each command, otherwise only the results of the last command will be saved in the default location.

You can overwrite this redirection by defining your own in the submitted command, e.g:

nuvolos apps execute -a your_app_slug -c 'python your_file.py > myoutput.log, 2> myerror.log'

that will save the stdout and stderr of your command to the defined files in the working directory.

Note

The Nuvolos API looks for the > redirection operator in the submitted command to determine whether custom redirection has been made. Keep in mind that using the operator in a different context in the command may result in unintended behavior.

Sharing results between applications

In the scenario where you've established a workflow with one application depending on the outcomes of another, the /files folder serves as a convenient storage space for files accessible to both applications, provided they reside within the same instance. If you wish to share results between applications located in separate instances, consider utilizing the Large File Storage feature on Nuvolos.