Skip to content
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

[Search] Inference Management FTR Tests #188438

Draft
wants to merge 13 commits into
base: main
Choose a base branch
from

Conversation

Samiul-TheSoccerFan
Copy link
Contributor

@Samiul-TheSoccerFan Samiul-TheSoccerFan commented Jul 16, 2024

Summary

FTR tests for Inference Management UI. Due to a race condition between license subscription and app registration, we are avoiding these tests for Stack only.

Checklist

Delete any items that are not applicable to this PR.

@Samiul-TheSoccerFan Samiul-TheSoccerFan requested review from a team as code owners July 16, 2024 13:24
@Samiul-TheSoccerFan Samiul-TheSoccerFan added release_note:skip Skip the PR/issue when compiling release notes Team:Search v8.15.0 v8.16.0 and removed v8.15.0 labels Jul 16, 2024
Copy link
Contributor

@TattdCodeMonkey TattdCodeMonkey left a comment

Choose a reason for hiding this comment

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

I still need to pull this down and test them, but if we're enabling the feature for ES3 then the tests should not be disabled in MKI IMO

const ml = getService('ml');

describe('Serverless Inference Management UI', function () {
this.tags('skipMKI');
Copy link
Contributor

Choose a reason for hiding this comment

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

if we're enabling the feature flag then we want to remove this tag. Enabling the feature means we WANT to run these against MKI.

We should verify the tests against MKI before merging this.

@@ -17,17 +17,17 @@ xpack.osquery.enabled: false
## Fine-tune the search solution feature privileges. Also, refer to `serverless.yml` for the project-agnostic overrides.
xpack.features.overrides:
### Dashboards feature is moved from Analytics category to the Search one.
dashboard.category: "enterpriseSearch"
dashboard.category: 'enterpriseSearch'
Copy link
Contributor

Choose a reason for hiding this comment

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

do we need these quote changes?

Copy link
Contributor Author

Choose a reason for hiding this comment

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

We do not need these quote changes. These go added when I rand lint command so I assumed something changed with the eslint configuration which forced the file to update.

@kibanamachine
Copy link
Contributor

Flaky Test Runner Stats

🎉 All tests passed! - kibana-flaky-test-suite-runner#6558

[✅] x-pack/test_serverless/functional/test_suites/search/config.ts: 25/25 tests passed.

see run history

@Samiul-TheSoccerFan Samiul-TheSoccerFan added the backport:skip This commit does not require backporting label Jul 16, 2024
@Samiul-TheSoccerFan
Copy link
Contributor Author

@elasticmachine merge upstream

@elasticmachine
Copy link
Contributor

elasticmachine commented Jul 16, 2024

💛 Build succeeded, but was flaky

  • Buildkite Build
  • Commit: 63c3b2a
  • Kibana Serverless Image: docker.elastic.co/kibana-ci/kibana-serverless:pr-188438-63c3b2a38df6

Failed CI Steps

Metrics [docs]

Async chunks

Total size of all lazy-loaded chunks that will be downloaded as the user navigates the app

id before after diff
searchInferenceEndpoints 84.7KB 85.3KB +591.0B

History

@Samiul-TheSoccerFan
Copy link
Contributor Author

Tests ran again MKI

Screenshot 2024-07-17 at 2 19 41 PM

@Samiul-TheSoccerFan
Copy link
Contributor Author

@TattdCodeMonkey : can you please give a review again?

@@ -65,7 +65,7 @@ data_visualizer.resultLinks.fileBeat.enabled: false
xpack.searchPlayground.ui.enabled: true

# Search InferenceEndpoints
xpack.searchInferenceEndpoints.ui.enabled: false
xpack.searchInferenceEndpoints.ui.enabled: true
Copy link
Contributor

@saikatsarkar056 saikatsarkar056 Jul 18, 2024

Choose a reason for hiding this comment

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

Based on the recommendation #5, we may need to hide the Inference Endpoints page in the servereless environment.

Copy link
Contributor Author

Choose a reason for hiding this comment

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

yes, we will not be merging this PR as the requirements changed

Copy link
Contributor

Choose a reason for hiding this comment

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

@Samiul-TheSoccerFan we can merge this PR, we just need to not enable the FF. instead run the new FTRs in the feature flag config and only enable it there.

num_threads: 1,
},
};
await ml.api.createInferenceEndpoint('endpoint-1', 'sparse_embedding', modelConfig);
Copy link
Contributor

@saikatsarkar056 saikatsarkar056 Jul 18, 2024

Choose a reason for hiding this comment

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

This endpoint was created in the ‘before’ block. We may need to write some cleanup code in the ‘after’ block to delete the endpoint and remove the underlying trained model if it was added during the creation of the endpoint.

I’m suggesting this because we previously had a CI/CD failure due to not cleaning up created resources. See this PR as an example of cleanup.

Copy link
Contributor Author

Choose a reason for hiding this comment

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

At the last test, we removed the inference endpoint from the UI. Do we still need to do the cleanup or it will be handled through the application itself?

Copy link
Contributor

Choose a reason for hiding this comment

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

Please check whether creation of inference endpoint adds any new trained model. If so, we need to clean up the trained model.

num_threads: 1,
},
};
await ml.api.createInferenceEndpoint('endpoint-1', 'sparse_embedding', modelConfig);
Copy link
Contributor

@saikatsarkar056 saikatsarkar056 Jul 18, 2024

Choose a reason for hiding this comment

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

Similar to this comment, we need to write some clean-up in the 'after' block.

@Samiul-TheSoccerFan
Copy link
Contributor Author

deprioritizing this PR as the plugin flow is expected to be changed. We will revisit this PR once we decide to move forward in Serverless.

@jeramysoucy
Copy link
Contributor

@Samiul-TheSoccerFan re:

deprioritizing this PR as the plugin flow is expected to be changed. We will revisit this PR once we decide to move forward in Serverless.

Could you switch this back to a draft PR?

@jeramysoucy jeramysoucy removed their request for review July 23, 2024 19:41
@Samiul-TheSoccerFan Samiul-TheSoccerFan marked this pull request as draft July 23, 2024 19:47
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
backport:skip This commit does not require backporting ci:build-serverless-image release_note:skip Skip the PR/issue when compiling release notes Team:Search v8.16.0
Projects
None yet
Development

Successfully merging this pull request may close these issues.

7 participants