-
Notifications
You must be signed in to change notification settings - Fork 1.2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
out: persitent outputs require odb when they should not #6533
Comments
@raylutz Workaround: setting up cache on
Reproduction script:
Seems like this issue is related to unprotect problem that one of our users have been having: EDIT: |
Thanks for the better name. I will tell you how I am now organizing this so I can avoid unnecessary transfers.
|
We have decided not to use DVC and have implemented our own similar functionality. Thanks for your time. |
Please note that using direct s3 paths as outputs/dependencies is an experimental scenario https://dvc.org/doc/user-guide/managing-external-data that has many considerations. We will be developing it properly in the future #3920 That being said, the bug here should be fixed, as it results in unnecessary |
Bug Report
Description
I am trying to use dvc to handle changes to program build data (not ML data). We don't need the cache and all data will be saved directly to s3. This is an unusual edge case and the Output.odb() property method assumes .repo.odb exists.
Reproduce
I have a single stage file:
The cmd probably can be just about anything, maybe 'echo "initialized"'
reference some s3: resources.
turn off caching
dvc init --no-scm
dvc repro -R -v --force dvc
Expected
Here is the error output I am getting:
Environment information
Output of
dvc doctor
:Additional Information (if any):
The text was updated successfully, but these errors were encountered: