Summary
The user suggests adding Python task execution on Databricks using a new Databricks agent with a different name, as it would lead to different behavior. They mention using Flytekit's entrypoint to run a Python function task on Databricks, but note that it is not necessary in this case. While combining everything is possible, the user expresses concern that the resulting code would be messy and lacks a clean, readable solution.
kumare
You don’t have to just wanted to see if anyone else wants it - cc <@U06AMB69K38> ?
eric901201
I can connect them, and ask them for getting some example
eric901201
does any user want that?
eric901201
yes I can do that
kumare
Absolutely different agent - but the same deployment secrets etc
eric901201
put them all together is doable, but the code will be super messy (or I didn't come out a solution to make code clean and readable for others
eric901201
but in your case, we don't have to use flytekit entrypoint
eric901201
since now we will use flytekit's entrypoint to run python function task on databricks
eric901201
different agent name will have different behavior
eric901201
with a different name in my opinion
eric901201
I think in this case, we have to create a new databrcisk agent
kumare
Shall we add python task execution on databricks through databricks agent?