We are getting a similar message when for several flows. This is a process that has worked for us for months. Some of the failing flows don't even exist in prefect cloud at all, so the complaint about a conflict is confusing. How can I troubleshoot?

A flow named 'pkt-anomaly-detection' and defined at '/builds/datateer/clients/pkt/pipeline/orchestration/pipeline_anomaly_detection.py:15' conflicts with another flow. Consider specifying a unique `name` parameter in the flow definition:

hi @Adam Roderick - is this a failure or just a warning? do you have more of the trace?

Normally this script results in just a new version of the deployment

the error isnt the name conflict though, thats just a warning

File "/builds/datateer/clients/dtr/pipeline/./datateer-prefect/datateer_prefect/release/infrastructure.py", line 108, in main
    run_shell_command(["prefect", "--no-prompt", "deploy"])
  File "/usr/local/lib/python3.9/site-packages/datateer_prefect/tasks/util.py", line 125, in run_shell_command
    raise ChildProcessError(os.linesep.join(errors)[:1000]) from None
ChildProcessError: Traceback (most recent call last):
raise self._exception
TypeError: 'NoneType' object is not iterable
An exception occurred.

this is the error

so that warning is not a problem?

no its not, its just saying that a flow with that name has been defined as some point in the past, you should be able to safely ignore that it seems like

errors

here is

NoneType

raise ChildProcessError(os.linesep.join(errors)[:1000]) from None

where

.join

wants it to be iterable

TypeError: 'NoneType' object is not iterable

Got it--I misread that it was a warning. Thank you @Nate

no problem!

The script is working for some flows but not others

im not sure what you've got in here

datateer-prefect/datateer_prefect/release/infrastructure.py

so I'm likely missing relevant context but in general yeah you need to tell

prefect deploy

which deployment you're referring to, unless you add

--all

at the end to deploy all flows in your

prefect.yaml

Is that new?

nope,

prefect deploy

either requires a

path/to/file.py:decorated_function

or

--name defined-in-prefect-yaml

Previously we ran a subprocess

prefect --no-prompt deploy

without specifying

--name

or

--all

glad to hear! sure thing