How to use Azure Batch with Prefetched Container from Custom Batch Service in Azure Data Factory

Question

Is it possible or is there documentation on how to run a batch service command from a pipeline in an Azure Data Factory against a batch pool that has a prefetched container image?

I created a batch pool that prefetches an image from our Azure container repository and it seems to start just fine. I can run simple a simple "hello world" command from the start task. However, every time I try to run a command from the Data Factory to this linked batch service, I get the following error:

Output
{
"exitcode": null,
"errorCategory": 0,
"code": "ContainerTaskSettingsNotFound",
"message": "Container-enabled compute node requires task container settings",
"details": null,

Is it possible to set TaskContainerSettings from the Data Factory? I see some doc here on doing this programmatically, but how can I get these JSON parameters to be created and sent from the Data Factory? (https://learn.microsoft.com/en-us/rest/api/batchservice/task/add#taskcontainersettings)

Answer 1

So I notice there is a section under 'Settings' called Extended Properties. This would be the place to supply any additional parameters.

As of May 15 2020, Data Factory did not support Containers on Azure Batch. At that time there were no plans to go beyond cluster operations.

@RichG