<#3588 [BUG] No Flyte Entities Detected with Ray e...
# flyte-github
a
#3588 [BUG] No Flyte Entities Detected with Ray example Jupyter Notebook workflow registration Issue created by peterklingelhofer Describe the bug Using the example Jupyter Notebook file found here: https://docs.flyte.org/projects/cookbook/en/latest/auto/integrations/kubernetes/ray_example/ray_example.html When I try to register the
ray_example.ipynb
file's workflow, I get this output:
Copy code
pyflyte register workflows --project repo --image localhost:30000/$FLYTE_DOCKER_IMAGE_TAG
Running pyflyte register from ~/Dev/repo with images ImageConfig(default_image=Image(name='default', fqn='localhost:30000/repo', tag='imageTag'), images=[Image(name='default', fqn='localhost:30000/repo', tag='imageTag')]) and image destination folder /root on 1 package(s) ('~/Dev/repo/workflows',)
Registering against localhost:30080
Detected Root ~/Dev/repo, using this to create deployable package...
No output path provided, using a temporary directory at /var/folders/t4/7xnpxvjj3w3g7cf752g1qvjh0000gn/T/tmp7dmep_68 instead
Computed version is kChHd1ibpkx7jpAR1MIWIg==
Loading packages ['workflows'] under source root ~/Dev/repo
Successfully serialized 0 flyte objects
No Flyte entities were detected. Aborting!
The
ray_example.ipynb
runs fine locally as we'd expect, it's just when attempting to register the workflow do we run into problems. Expected behavior I would expect this to work the way the python example (downloadable from that same page in the docs) works:
Copy code
pyflyte register workflows --project repo --image localhost:30000/$FLYTE_DOCKER_IMAGE_TAG
Running pyflyte register from ~/Dev/repo with images ImageConfig(default_image=Image(name='default', fqn='localhost:30000/repo', tag='imageTag'), images=[Image(name='default', fqn='localhost:30000/repo', tag='imageTag')]) and image destination folder /root on 1 package(s) ('~/Dev/repo/workflows',)
Registering against localhost:30080
Detected Root ~/Dev/repo, using this to create deployable package...
No output path provided, using a temporary directory at /var/folders/t4/7xnpxvjj3w3g7cf752g1qvjh0000gn/T/tmpq6mm2c8o instead
Computed version is 69UqeNHQfi0GyndHMSA6uw==
Loading packages ['workflows'] under source root ~/Dev/repo
Successfully serialized 3 flyte objects
[✔] Registration workflows.ray_example.ray_task type TASK successful with version 69UqeNHQfi0GyndHMSA6uw==
[✔] Registration workflows.ray_example.ray_workflow type WORKFLOW successful with version 69UqeNHQfi0GyndHMSA6uw==
[✔] Registration workflows.ray_example.ray_workflow type LAUNCH_PLAN successful with version 69UqeNHQfi0GyndHMSA6uw==
Successfully registered 3 entities
Additional context to reproduce I noticed that VS Code thinks there is a
\n
after
@workflow
(unsurprising as Jupyter Notebooks are typically run in the browser obviously), not sure if that could be causing the problem. Screenshots

Screenshot 2023-04-12 at 3 47 20 PM

Are you sure this issue hasn't been raised already? ☑︎ Yes Have you read the Code of Conduct? ☑︎ Yes flyteorg/flyte