fix: remove historical job_name caching which causes long job name

[jkasiraj]

Issue #, if available: #5099

Description of changes:

• Remove historical job_name caching which causes job name to exceed 63 characters with the addition of the 64 char sha256 hash
• Both TrainingJobName and ProcessingJobName have character limits of 63:
  • https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_CreateProcessingJob.html#sagemaker-CreateProcessingJob-request-ProcessingJobName
  • https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_CreateTrainingJob.html#sagemaker-CreateTrainingJob-request-TrainingJobName
• local artifact hash is now used instead of timestamp in the s3 path to support caching usecases

Testing done:

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 backward 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)
• If adding any dependency in requirements.txt files, I have spell checked and ensured they exist in PyPi

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

[benieric]

Can we generate the name using the name of the image? That is what other interfaces do. Does removing these code blocks break anything? Maybe need a closer look but if 'job_name = None' we may need to generate a name some other way no?

[benieric]

https://github.com/aws/sagemaker-python-sdk/blob/master/src/sagemaker/pipeline.py#L244
https://github.com/aws/sagemaker-python-sdk/blob/master/src/sagemaker/modules/train/model_trainer.py#L452

[jkasiraj]

Sure I think we can set it to self.name, since the removed code was setting it to f"{self.name}-{code_hash}"[:1024]

I believe right now the behavior is that if an estimator is set in a particular way (logic from lines 497-506) it leads to a job name that's too long + errors or otherwise it defaults to None, which is why I considered leaving it as None

I'm a bit hesitant to add new behavior like trying to parse the image (which technically isn't required by the api)

• https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_CreateTrainingJob.html

[jkasiraj]

Actually, I looked into it further and the job name will eventually get generated here if it doesn't exist so I don't think it needs to be set in this function

[benieric]

Gotchu sounds good 👍