Defining commands and arguments for your workloads

When you create a container image for your IBM Cloud® Code Engine workloads, you can define commands and arguments for your job or application to use at run time.

Container images include two pieces of metadata that tell the container runtime what command, inside the image to run when the container is created. These metadata fields are called Entrypoint and Command. For those users who are familiar with Dockerfile, the fields equate to the ENTRYPOINT and CMD commands. These two fields contain arrays of strings that are combined to create the command line that is used when you run your container.

For example, if your container image has an Entrypoint value of /myapp and a Command value of --debug, then the full command that is run is /myapp --debug (an array of two strings). Notice that because this action is a concatenation of two arrays, if Entrypoint is an empty array then the Command array's first array element is the executable that is run in your container.

When you create a Code Engine application or job, you can provide values for both the Entrypoint and Command arrays.

• If --command is used, then any image Entrypoint value is overwritten and any image Command value is ignored.
• If --argument is used, then any image Command value is overwritten.

To better understand this process, look at the following examples,

You can specify these values by using the --command and --argument options in the CLI for apps and jobs and the Command and Arguments entry boxes in the console for jobs. For more information, see Deploying your app with commands and arguments and Running your job with commands and arguments.