Skip to content

fix: TrainingStep cache misses due to timestamp based job name #3070

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.

Sign up for GitHub

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

Jump to bottom
Merged
merged 1 commit into from
Apr 19, 2022
Merged

fix: TrainingStep cache misses due to timestamp based job name #3070

merged 1 commit into from
Apr 19, 2022

Conversation

nmadan
Copy link
Member

@nmadan nmadan commented Apr 15, 2022

Issue #, if available:

Description of changes: Present cache misses for TrainingStep reruns by changing the training job name, as well as upload path of user's training scripts. New name is calculated by hashing the inputted scripts/source_dir/dependencies. This change is similar to #2790

Testing done:
unit testing

Merge Checklist

Put an x in the boxes that apply. You can also fill these out after creating the PR. If you're unsure about any of them, don't hesitate to ask. We're here to help! This is simply a reminder of what we are going to look for before merging your pull request.

General

  • I have read the CONTRIBUTING doc
  • I certify that the changes I am introducing will be backword compatible, and I have discussed concerns about this, if any, with the Python SDK team
  • I used the commit message format described in CONTRIBUTING
  • I have passed the region in to all S3 and STS clients that I've initialized as part of this change.
  • I have updated any necessary documentation, including READMEs and API docs (if appropriate)

Tests

  • I have added tests that prove my fix is effective or that my feature works (if appropriate)
  • I have added unit and/or integration tests as appropriate to ensure backward compatibility of the changes
  • I have checked that my tests are not configured for a specific region or account (if appropriate)
  • I have used unique_name_from_base to create resource names in integ tests (if appropriate)

By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.

@nmadan nmadan requested a review from staubhp April 15, 2022 01:04
@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-unit-tests
  • Commit ID: d941489
  • Result: FAILED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-local-mode-tests
  • Commit ID: d941489
  • Result: SUCCEEDED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-unit-tests
  • Commit ID: d53aad6
  • Result: FAILED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-unit-tests
  • Commit ID: 4da8a3d
  • Result: FAILED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-unit-tests
  • Commit ID: dc793f8
  • Result: FAILED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-unit-tests
  • Commit ID: 63e52f8
  • Result: FAILED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-local-mode-tests
  • Commit ID: 4da8a3d
  • Result: SUCCEEDED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-local-mode-tests
  • Commit ID: d53aad6
  • Result: SUCCEEDED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@codecov-commenter
Copy link

codecov-commenter commented Apr 15, 2022 •
edited
Loading

Codecov Report

Merging #3070 (405d1bf) into master (fd753c3) will increase coverage by 0.00%.
The diff coverage is 92.15%.

@@           Coverage Diff           @@
##           master    #3070   +/-   ##
=======================================
  Coverage   89.38%   89.38%           
=======================================
  Files         197      197           
  Lines       16670    16715   +45     
=======================================
+ Hits        14900    14941   +41     
- Misses       1770     1774    +4
Impacted Files Coverage Δ
src/sagemaker/workflow/utilities.py 92.85% <87.87%> (-7.15%) ⬇️
src/sagemaker/estimator.py 91.30% <100.00%> (ø)
src/sagemaker/workflow/steps.py 97.95% <100.00%> (+0.14%) ⬆️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update fd753c3...405d1bf. Read the comment docs.

@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-unit-tests
  • Commit ID: 9deb2f3
  • Result: SUCCEEDED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-local-mode-tests
  • Commit ID: dc793f8
  • Result: SUCCEEDED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-local-mode-tests
  • Commit ID: 63e52f8
  • Result: SUCCEEDED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-slow-tests
  • Commit ID: 63e52f8
  • Result: SUCCEEDED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-slow-tests
  • Commit ID: dc793f8
  • Result: FAILED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

mufaddal-rohawala
mufaddal-rohawala reviewed Apr 15, 2022
View reviewed changes
src/sagemaker/workflow/utilities.py Outdated
"""
if isinstance(file, str) and file.lower().startswith("file://"):
file = unquote(urlparse(file).path)
assert Path(file).is_file()
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can we raise a more specific error here and other places instead of assertion.

@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-unit-tests
  • Commit ID: 405d1bf
  • Result: SUCCEEDED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-unit-tests
  • Commit ID: e84d37c
  • Result: SUCCEEDED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-local-mode-tests
  • Commit ID: e84d37c
  • Result: FAILED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-local-mode-tests
  • Commit ID: 405d1bf
  • Result: SUCCEEDED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-slow-tests
  • Commit ID: 405d1bf
  • Result: FAILED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

@sagemaker-bot
Copy link
Collaborator

AWS CodeBuild CI Report

  • CodeBuild project: sagemaker-python-sdk-slow-tests
  • Commit ID: e84d37c
  • Result: FAILED
  • Build Logs (available for 30 days)

Powered by github-codebuild-logs, available on the AWS Serverless Application Repository

jerrypeng7773
jerrypeng7773 reviewed Apr 19, 2022
View reviewed changes
src/sagemaker/workflow/steps.py
# To avoid this, hash the contents of the training script and include it
# in the `job_name` passed to the `Estimator`, which will be used
# instead of the timestamped path.
self.job_name = self._generate_code_upload_path()
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

but job_name later will be excluded from the definition file right? how the backend decide when to use cache or not tho?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Backend uses a hash of all the input arguments to the step + pipeline name + step name. The issue was that the bucket name where all the local scripts were uploaded had a timestamp from when the pipeline was run. This bucket name is passed in as a hyperparameter to the training job.

@ahsan-z-khan ahsan-z-khan changed the title fix: fix TrainingStep cache misses due to timestamp based job name fix: TrainingStep cache misses due to timestamp based job name Apr 19, 2022
@ahsan-z-khan ahsan-z-khan merged commit a3cb74b into aws:master Apr 19, 2022
jerrypeng7773 pushed a commit to jerrypeng7773/sagemaker-python-sdk that referenced this pull request May 13, 2022
@nmadan @jerrypeng7773
fix: TrainingStep cache misses due to timestamp based job name (aws#3070
da906dc 
)

Co-authored-by: ci <ci>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@navinsoni navinsoni navinsoni left review comments

@staubhp staubhp staubhp approved these changes

@ahsan-z-khan ahsan-z-khan ahsan-z-khan approved these changes

@jerrypeng7773 jerrypeng7773 jerrypeng7773 approved these changes

@mufaddal-rohawala mufaddal-rohawala mufaddal-rohawala approved these changes

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
8 participants
@nmadan @sagemaker-bot @codecov-commenter @navinsoni @staubhp @ahsan-z-khan @jerrypeng7773 @mufaddal-rohawala