Template deployment creates an append transaction for output dataset

Ask the Community
1

Hey team,
I know template is sunset but we are still using it in my team because the migration to marketplace is quite complex. I was wondering the conditions for a template deployment to create an append transaction for the output dataset of a transform? I can see in prod that the template deployment always create an append transaction with 0 bytes of data and some custom metadata

{
  "spark-catalog": {
    "master": {
      "partitionColumnNames": [],
      "bucketSpec": null,
      "partitions": []
    }
  }
}

When I try and replicate in a test environment, I only have an append transaction on first deployment, and I don’t get it on second deployment anymore even after changing the transform. Does anyone know why there is an append transaction in the first place and the conditions for a template deployment to trigger one?

Thanks!

2

Hi, just to clarify, is the issue that you expect on the second deployment, for the transform to build a second time and for there to be another append transaction or what is the expectation?

3

Hi Emily, the expectation is what I observe in production - anytime a template is deployed, an append transaction is created on the output datasets by template, regardless of whether or not you build the transform