Image Builder Improvements #500
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
- refactor pyfunc base image
added 28 commits
November 30, 2023 19:53
…e sure this layer got cached
… env value is always defined hence the next run will not invalidated
krithika369
reviewed
Dec 7, 2023
leonlnj
reviewed
Dec 7, 2023
leonlnj
reviewed
Dec 7, 2023
|
-> Why is the python/sdk/test/pyfunc/model_2.joblib file removed in this PR? Isn't it needed by the integration tests? When running pyfunc_integration_test.py from local machine, it will re-train the model hence updating the model. I deleted it but forgot to restore the old one. I've restored it know. |
krithika369
reviewed
Dec 8, 2023
krithika369
reviewed
Dec 8, 2023
krithika369
reviewed
Dec 8, 2023
krithika369
reviewed
Dec 8, 2023
There was a problem hiding this comment.
@ariefrahmansyah the PR largely LGTM (need to remember to revert some of the TODO in the GitHub workflows before merging). However, I left some new comments; especially around the reuse of MLP's artifact.Service instead of gsutil which I believe would be a lot simpler. Will re-review this part once it's addressed. Thanks!
added 3 commits
December 8, 2023 14:16
2. Add SupportedPythonVersions field 3. Refactor BuildContextSubPath to be part of BaseImageConfig
2 tasks
krithika369
approved these changes
Dec 10, 2023
There was a problem hiding this comment.
Left a couple of small comments. The rest LGTM. Thanks for this PR and the CI steps refactoring, @ariefrahmansyah !
5 tasks
ariefrahmansyah
added a commit
that referenced
this pull request
Dec 15, 2023
<!-- Thanks for sending a pull request! Here are some tips for you: 1. Run unit tests and ensure that they are passing 2. If your change introduces any API changes, make sure to update the e2e tests 3. Make sure documentation is updated for your PR! --> **What this PR does / why we need it**: <!-- Explain here the context and why you're making the change. What is the problem you're trying to solve. ---> For existing models that were previously successfully built using Python 3.7.* won't be able to be deployed anymore since the new validatePythonVersion() introduced in #500 will invalidate it. **Which issue(s) this PR fixes**: <!-- *Automatically closes linked issue when PR is merged. Usage: `Fixes #<issue number>`, or `Fixes (paste link of issue)`. --> Fixes failed redeployment or re-execution of existing models with Python 3.7.* **Does this PR introduce a user-facing change?**: <!-- If no, just write "NONE" in the release-note block below. If yes, a release note is required. Enter your extended release note in the block below. If the PR requires additional action from users switching to the new release, include the string "action required". For more information about release notes, see kubernetes' guide here: http://git.k8s.io/community/contributors/guide/release-notes.md --> ```release-note NONE ``` **Checklist** - [ ] Added unit test, integration, and/or e2e tests - [x] Tested locally - [ ] Updated documentation - [ ] Update Swagger spec if the PR introduce API changes - [ ] Regenerated Golang and Python client if the PR introduce API changes
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What this PR does / why we need it:
Image building is a mandatory process in the deployment of PyFunc-based Merlin components. The process abstracts the complexity of building and publishing a ready-to-be-used Docker image that contains the user's model source code, artifacts, and dependencies. However, there are some issues with the current image-building approach that lead to poor development experience and inefficient resource utilization. This PR addresses them by:
a. it is more configurable -- we can set Kaniko's build arguments flags in the config and they will be propagated to the kaniko job spec
b. it supports caching
a. To achieve it, we introduce a new gsutil package to interface with GCS
a. Previously, the base Dockerfile had a step to install the pyfunc-server or batch prediction jobs dependencies. This step is now executed by the application Dockerfile
i. Subsequently, this allows us to have only one base image instead of building base images for each Python version.
b. New base Dockerfile will only install the platform components such as conda, spark, grpc library, but not the component itself
c. New application Dockerfile will first download and install user's model dependencies; install component's dependencies; download model source code and artifacts; and finally dry-run the model
Which issue(s) this PR fixes:
Improves image building performance
Does this PR introduce a user-facing change?:
Checklist