Hi All, I’m having a strange problem executing SQL command within tasks in a Flow. To summerise, I’m using sqlalchemy to execute those commands for me on the database. the issue is that commands are getting executed with no error returned, but later looking at the database, none of those commands went into effect (as if the were not executed at all). Surprsingly, another task within the same Flow is able to successfully utilize the engine (instantiated the same way), when pandas is using that to dump a dataframe into the database. I did a test on the docker environment, where the tasks are executed, and surprisigly, all those sql command are getting executed successfully. I’m a little puzzled as to why a task can utilize sqlalchemy engine and another task fails to do so? please help.

Hey @Mehdi Nazari, can I see the code? Maybe it’s just the return? Like the cursor is not executing?

Hey @Kevin Kho, this is the stripped down version after a few adjustment to make sure command are valid.

This looks very similar to the Prefect tasks. I don’t see anything wrong. And I guess you said it worked in the Docker environment with the same code? I guess it would be a connection issue then?

It did actually, on the same docker container, I tested this code in a python REPL to make sure that environment doesn’t have any problem. they were all successful.

I don’t see anything wrong. Why don’t you try

task.run()

so run the task alone without a Flow? And then do you see the logs there? Like the info one.

This is a task right before the one above; I’m surprised this succeedes

Just execute it like python code

load_to_mssql.run()

. This will run the Python code underneath

So strange!

task.run()

executes successfully from my laptop! does that tell you something?

It doesn’t… cuz that’s just what the Flow does! If you see the task running and you see the logs when you run the Flow,

flow.run()

and running the Flow are doing the same thing. 🤔

I did

python flow.py

that has a register command in there, also added the

task.run()

before registering. I’m assuming that run in locally on my machine with no agent involved?

Yes that just calls the Python function underneath but before

task.run()

, how did you run it?

Just like I mentioned,

python flow.py

ran it for me. and it was successful.

Yeah!

@task()
def abc(a):
    return a
with Flow("ecs_testing") as flow:
    x = abc(1)
    y = abc(1, upstream_tasks=[x])
with Flow("ecs_tes") as flow2:
    x = abc(1)
    y = abc(1)
    y.set_upstream(x)