Snowpark Migration Accelerator: Interpreting the Assessment Output¶

When the tool has finished running, the analysis is complete. So now what? What decisions can be made from the output? How can you get value out of all of this data? Let’s look at the output.

When the analysis is complete, a summary will be shown in the application. The complete assessment information will be populated in the output directory specified when the project was created. The artifacts created in this directory will be where you find the majority of value from the assessment.

This walkthrough will review how to get value out of both the in application output, and the complete output directory, but what we’re looking for is summarized here:

• Readiness Score(s) - The SMA produces multiple readiness scores designed to give multiple dimensions of readiness. This “readiness” could be described as compatibility between this codebase and Snowflake. For example, the Spark API Readiness Score shows how many of the references to the Spark API present in the source code are ready to be converted to the Snowpark API. The SQL Readiness Score shows how many of the Spark SQL (or HiveQL) references in the source code are ready to be converted to Snowflake SQL. These are the key indicators to better understand the compatibility of this workload with Snowflake, and an estimate on how much of the code will need to be manually converted.

• Size - Size can be evaluated by using the File Summary. The size of the workload in both file count and lines of code allows you to understand how large this workload will have. Combined with the readiness score, you can understand the readiness level by file to determine which files are “ready” for Snowpark. This allows you to determine if there is a subset of files that have a higher or lower readiness and if those will require more attention.

• More to come, but in the meantime…

Let’s start by reviewing the summary page in the application.