-
Notifications
You must be signed in to change notification settings - Fork 996
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
Better logging for materialize command #1499
Better logging for materialize command #1499
Conversation
sdk/python/feast/infra/gcp.py
Outdated
lambda b: _write_minibatch(client, project, table, b, progress), | ||
_to_minibatches(data), | ||
) | ||
with tqdm(total=len(data)) as pbar: |
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.
I think it may be better to have tqdm at a higher level so that its not optional to do progress bar logging. Would it make more sense in materialize_incremental()
and then pass it into materialize_single_feature_view
?
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.
I don't see an easy way to bring it up to materialize_incremental since we don't know the number of rows of data at that level; however, progress
is an argument in methods in the Provider
class so that kinda makes it not optional for implementations of Provider
to do some sort of progress logging, presumably by copying existing implementations and using tqdm
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.
I thought I replied here, but somehow my message didn't send. I think my question is just: How can we ensure consistency in the logging output if the implementations need to create their own tqdm object. Wouldn't the current implementation make it unscalable to have different configurations for tqdm? Even if the row count is deeper, maybe it makes sense to instantiate the tqdm object higher and pass it into the implementation.
Nitpicks:
edit: For none of these points I feel super strongly though. Happy to keep as is. |
/kind housekeeping |
Signed-off-by: Jacob Klegar <jacob@tecton.ai>
Signed-off-by: Jacob Klegar <jacob@tecton.ai>
Signed-off-by: Jacob Klegar <jacob@tecton.ai>
aa2757c
to
0ebf141
Compare
/lgtm |
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: jklegar, tsotnet The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
* Better logging for materialize command Signed-off-by: Jacob Klegar <jacob@tecton.ai> * Address comments Signed-off-by: Jacob Klegar <jacob@tecton.ai> * Move tqdm to FeatureStore class Signed-off-by: Jacob Klegar <jacob@tecton.ai> * Rebase Signed-off-by: Jacob Klegar <jacob@tecton.ai>
* Better logging for materialize command Signed-off-by: Jacob Klegar <jacob@tecton.ai> * Address comments Signed-off-by: Jacob Klegar <jacob@tecton.ai> * Move tqdm to FeatureStore class Signed-off-by: Jacob Klegar <jacob@tecton.ai> * Rebase Signed-off-by: Jacob Klegar <jacob@tecton.ai>
Signed-off-by: Jacob Klegar jacob@tecton.ai
What this PR does / why we need it: Improve logging for materialize and materialize_incremental
Which issue(s) this PR fixes:
Fixes #
Does this PR introduce a user-facing change?: