-
Notifications
You must be signed in to change notification settings - Fork 5
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
bugfix: DAG import error for TaskGroup
#35
Conversation
Signed-off-by: Fredrik Bakken <[email protected]>
Hi @FredrikBakken, thanks for your contribution! Happy to have you on-board! |
Hi @FredrikBakken ! Thanks for this PR, it's much appreciated. I am by no means expert in Airflow, but this confuses me (and is probably why we have this assumption in the code):
Can you share some context from your calling code, please? Also, would it be acceptable to you to raise an error if |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hi again @FredrikBakken,
Following #36 and its fix in #37, it turns out that currently our CI fails on this PR. Could you please rebase on top of it so we can test? (I cannot do it in practice because it's on a fork -- so if you cannot I shall merely do so manually...).
@FredrikBakken: once rebased (#38), this passes. 🎇 Gentle reminder: The blocker to merging is this comment regarding its necessity in Airflow DAGs. I understand from the docs close to the exact opposite -- that every task must have a task_id. If you could point me at using code or an example from other tasks in other providers, that would be very helpful! Thanks again :-) |
Closing due to lack of traction. @FredrikBakken thanks for your work on this! Of course e will happily reopen when you have more time for this. |
@arielshaqed, sorry for the late response to this PR. It has been a couple of busy months and I have not had time to keep up to date on everything. I am now back to using LakeFS (with Airflow) on some local tests and have re-encountered this issue again on My Airflow task looks as follows: task_fetch_dataset = LakeFSUploadOperator(
task_id='upload_file',
repo=default_args.get('repo_bronze'),
branch=f'ssb_dataset_{date}',
path='dataset',
content=NamedStringIO(
content="",
name='dataset.json')) I agree that I also took a look at the official BashOperator from Airflow, which does not define Let me know if there are anything you want me to change on the PR 🤝 |
Sorry for not addressing this sooner -- I failed to see the comment, because I don't regularly inspect closed PRs. Which is particularly unpleasant given that I closed the PR. I went over your explanations and now understand better. I will try to rerun CI on this or on a copy, and then see if we can get this fix in. |
No worries, @arielshaqed ! Great to hear, and please let me know if there are any things I can do on my side. |
TaskGroup
@arielshaqed, I've updated the title of this PR to better define the scope of what it solves. Using the operators defined in the #34 table works well in single task mode, but once the tasks are placed within a |
97ea5c3
to
da2e8e6
Compare
@arielshaqed, should I add some provider tests to this PR to confirm that the following operators, |
Sorry. Opened #77 for this, I've blocked this PR for too long already. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
THANKS!
For further explanation about this bugfix, please see #34.
Closes #34, Closes #44, Closes #69