Fix reusable k8s training service bug #5045
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
…ero/nni into env_k8s_trainingservice_fix
|
|
|
The commit(0c9c151) has a side effect when the storage is not nfs(like azure). To be refined. |
liuzhe-lz
approved these changes
Aug 12, 2022
ultmaster
approved these changes
Aug 15, 2022
|
Need to be tested on K8S pipline. Tracked in #4954. |
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Outdated suggestions cannot be applied.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.
Description
For now, the reusable k8s training service(framework controller and kubeflow) will save the entry point file(run.sh) and env dir at the root path of NFS, but the content of run.sh declare it should be located at "envs/exp_id". So, the "cannot find xxxx" error will be raised when the worker pod runs this entry point.
Second, when users submit multi-experiments, different experiments will affect each other(due to all contents are stored at the root of NFS).
And the entry-point will be overwritten by different environments when the trialConcurrency is more than 1, so each environment will actually run the same(the latest generated) run.sh script, and https://github.com/microsoft/nni/blob/v2.8/nni/tools/trial_tool/trial_runner.py#L164 will use dir name as runner id, finally, it will raise an error.
More discussion can be found at #4588
Test Options
Checklist
How to test
Related Issues: microsoft/frameworkcontroller#75, #4588, #4874, #5026.