Hello there! :wave: Hope this is the right place t...
# ask-the-community
a
Hello there! 👋 Hope this is the right place to ask I just started testing out Flyte. Following this tutorial, I keep getting some FileNotFound errors when I execute the workflow remotely on the sandbox cluster (
pyflyte run --remote filebased_wf.py wf --indirpath input-data
), despite creating the files in minio from browser. On the contrary, I can execute workflows which write files (from tmp). Is there any restriction on reading from minio in sandbox? How should I do it?
For reference, the updated tutorials on FlyteFiles I found in the doc don't read files from s3, rather download them from url - that's also where the question stems from
k
Pyflyte should automatically upload files if you point to local files
Don’t need to upload manually to minio
a
Thanks Ketan! My point was to try out a flow in which my workflow is reading data from S3 as later I'll need to read from GCS. I see that's not needed if I have local files, but what if I don't? Is there anything particular I should do to read from storage?
s
as long as flyte has access to the blob storage, it should work. did you provide s3 uri as input to the workflow?
a
Thanks Samhita! I was providing with the bucket folder but not the URI, that solved accessing the files! Thank you