hey guys I have a flow in prefect cloud that’s sho...
# prefect-community
k
hey guys I have a flow in prefect cloud that’s showing this error
{'_schema': 'Invalid data type: None'}
after it is scheduled but before the first task gets executed. When I try to run this flow in a local environment it executes as expected. Any idea what could be causing this issue?
k
I am wondering if this might be a version mismatch between agent/flow registration/runtime? You’re on Cloud right?
k
yep
k
This guy had the same issue and it was an old agent
k
what’s weird is i’m using the same agent to register a lot of other flows in the same cloud environment on a different project and the other flows seem to be executing fine
k
Are they all the same storage and runconfig?
k
yea
k
Are you using something with image maybe? Like ECSRun that is pulling a new Prefect image?
k
no it’s just a simple etl, querying from snowflake and writing to psql
my executor is a local dask executor too if that helps for context
k
Ah I can’t pinpoint what causes this error, but it is good practice though to have agent version higher than flow version
k
got it I will see if i can troubleshoot and update here
huh so i reordered some of the python code that we are using for flow execution and it seems you can’t add a flow to a Docker storage from
prefect.storage.Docker
using
add_flow()
if that storage variable has already been set on a flow ex:
flow.storage = storage
reordering the python execution order seems to have fixed the problem. The flow runs in prefect cloud now.
k
blinking
that’s news to me but glad you got it working