MAF-19283: feat(e2e): add Heimdall tag support and enhance inference service configurations · Pull Request #56 · moreh-dev/mif

ghost · 2026-02-12T10:46:00Z

No description provided.

…service configurations - Introduced a new environment variable `HEIMDALL_TAG` for Heimdall configuration. - Updated inference service data structure to include `Replicas` and `HeimdallTag`. - Modified performance and quality test files to utilize the new `HeimdallTag` variable. - Enhanced YAML templates for Heimdall to dynamically set the image tag based on the new variable. - Improved resource management in inference service configurations for better clarity and maintainability.

…lude HEIMDALL_TAG environment variable for enhanced configuration management

…del PV and PVC management. Moved model PV and PVC creation to a conditional block based on the SkipKind environment variable, improving resource handling. Streamlined the deletion of Heimdall to ensure clarity in resource management during test teardown.

Copilot

Pull request overview

This PR enhances the E2E performance/quality suites by allowing the Heimdall image tag to be configured via an environment variable and by making InferenceService manifests more configurable (notably replica counts and non-kind scheduling constraints).

Changes:

Add HEIMDALL_TAG env var support and pass it into Heimdall Helm values templates.
Parameterize InferenceService spec.replicas via InferenceServiceData.Replicas and update tests accordingly.
Extend performance InferenceService configuration with additional node affinity/tolerations, and add extra volumes/mounts in performance Heimdall values.

Reviewed changes

Copilot reviewed 9 out of 9 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
`test/utils/inference_service.go`	Adds `Replicas` to template render data and removes the now-unused helper constructor.
`test/e2e/quality/quality_test.go`	Passes Heimdall tag to values template and sets replicas explicitly for the quality InferenceService.
`test/e2e/quality/config/inference-service.yaml.tmpl`	Replaces hardcoded replicas with templated `.Replicas`.
`test/e2e/quality/config/heimdall-values.yaml.tmpl`	Introduces `image.tag` templating via `.HeimdallTag`.
`test/e2e/performance/performance_test.go`	Passes Heimdall tag to values template and sets replicas explicitly for prefill/decode services.
`test/e2e/performance/config/inference-service.yaml.tmpl`	Replaces hardcoded replicas with `.Replicas` and adds non-kind affinity/tolerations.
`test/e2e/performance/config/heimdall-values.yaml.tmpl`	Adds `image.tag` templating and configures extra volumes/mounts.
`test/e2e/envs/env_vars.go`	Registers `HEIMDALL_TAG` in centralized E2E env var management.
`.github/workflows/e2e-inference-perf-p-cluster.yaml`	Sets `HEIMDALL_TAG` for the performance workflow run.

…e YAML template to enhance resource scheduling

ghost · 2026-02-12T11:11:13Z

@hyeongyun0916 지금 작업사항대로 Action 실행하면 정상 작동합니다.

https://github.com/moreh-dev/mif/actions/runs/21943807909

inference-perf를 실행하고 나면 실험 결과를 s3에 저장하고 바로 namespace 통째로 리소스를 정리해 버리므로,
결과를 확인하시려면 아래 두 가지 방법 중 하나를 선택하시면 됩니다.

Action의 환경변수 설정에 SKIP_CLEANUP: "true" 추가하여 리소스 정리되지 않게 한 다음 namespace 접속해서 pod log 확인
aws s3에 저장된 파일을 가져와서 확인
i. aws s3 ls --profile s3 s3://moreh-benchmark/ --recursive

Copilot

Pull request overview

Copilot reviewed 9 out of 9 changed files in this pull request and generated 1 comment.

hyeongyun0916 · 2026-02-13T00:38:28Z

@hyeongyun0916 지금 작업사항대로 Action 실행하면 정상 작동합니다.

https://github.com/moreh-dev/mif/actions/runs/21943807909

inference-perf를 실행하고 나면 실험 결과를 s3에 저장하고 바로 namespace 통째로 리소스를 정리해 버리므로, 결과를 확인하시려면 아래 두 가지 방법 중 하나를 선택하시면 됩니다.

Action의 환경변수 설정에 SKIP_CLEANUP: "true" 추가하여 리소스 정리되지 않게 한 다음 namespace 접속해서 pod log 확인

aws s3에 저장된 파일을 가져와서 확인
i. aws s3 ls --profile s3 s3://moreh-benchmark/ --recursive

이거 md나 action파일위쪽에 주석이나 괜찮은곳으로 추가부탁드립니다.

hhk7734

heimdall tag -> heimdall version
image.tag 명시 제거
heimdall helm version을 변경

…o include result retrieval instructions for S3, enhancing clarity on post-test result management.

ghost · 2026-02-13T01:27:15Z

heimdall tag -> heimdall version
image.tag 명시 제거
heimdall helm version을 변경

"테스트는 마이너 버전을 release해서 진행하던 무조건 버전값이 있어야 하고, tag를 변경해서 진행하는건 안된다" 라고 이해하면 될까요?

heimdall version은 여기서 변경할 수 있습니다
https://github.com/moreh-dev/mif/blob/main/test/utils/settings/constants.go#L54

…ment variables and test configurations to streamline setup and enhance clarity in performance and quality tests.

ghost · 2026-02-13T01:29:37Z

이거 md나 action파일위쪽에 주석이나 괜찮은곳으로 추가부탁드립니다.

b35ae98

ghost · 2026-02-13T01:29:48Z

heimdall tag -> heimdall version image.tag 명시 제거 heimdall helm version을 변경

ab61b26

hhk7734 · 2026-02-13T01:32:47Z

heimdall tag -> heimdall version
image.tag 명시 제거
heimdall helm version을 변경

"테스트는 마이너 버전을 release해서 진행하던 무조건 버전값이 있어야 하고, tag를 변경해서 진행하는건 안된다" 라고 이해하면 될까요?

heimdall version은 여기서 변경할 수 있습니다

https://github.com/moreh-dev/mif/blob/main/test/utils/settings/constants.go#L54

vX.Y.Z, vX.Y.Z-rc.N 외의 테스트는 E2E에서 다루지 않겠습니다.

Copilot

Pull request overview

Copilot reviewed 7 out of 7 changed files in this pull request and generated no new comments.

…tests by introducing a conditional check for the IsKind variable, improving clarity in resource management and template rendering based on the environment configuration.

…ope formats to improve testing guidelines

ghost · 2026-02-13T01:38:26Z

vX.Y.Z, vX.Y.Z-rc.N 외의 테스트는 E2E에서 다루지 않겠습니다.

이 부분 기록 차원으로 AGENTS.md에 추가하겠습니다
40e5bcb

Copilot

Pull request overview

Copilot reviewed 8 out of 8 changed files in this pull request and generated no new comments.

seongsu-dev added 2 commits February 12, 2026 19:39

MAF-19283: feat(e2e): Update e2e-inference-perf-p-cluster.yaml to inc…

0505bb9

…lude HEIMDALL_TAG environment variable for enhanced configuration management

ghost self-assigned this Feb 12, 2026

ghost self-requested a review as a code owner February 12, 2026 10:46

ghost requested review from TaeSooRhee-moreh, Copilot and nulledge February 12, 2026 10:46

Copilot started reviewing on behalf of ghost February 12, 2026 10:46 View session

Copilot AI reviewed Feb 12, 2026

View reviewed changes

Comment thread test/e2e/performance/config/heimdall-values.yaml.tmpl Outdated

Comment thread test/e2e/quality/config/heimdall-values.yaml.tmpl Outdated

Comment thread test/e2e/performance/config/heimdall-values.yaml.tmpl

Comment thread test/e2e/performance/config/inference-service.yaml.tmpl

MAF-19283: feat(e2e): Add toleration for AMD GPUs in inference servic…

5883e3a

…e YAML template to enhance resource scheduling

ghost marked this pull request as draft February 12, 2026 10:54

ghost requested a review from hyeongyun0916 February 12, 2026 11:02

Merge branch 'main' into MAF-19283_add_toleration_affinity

3c59d57

ghost marked this pull request as ready for review February 12, 2026 11:17

Copilot AI review requested due to automatic review settings February 12, 2026 11:17

Copilot started reviewing on behalf of ghost February 12, 2026 11:17 View session

Copilot AI reviewed Feb 12, 2026

View reviewed changes

Comment thread test/e2e/performance/config/heimdall-values.yaml.tmpl

hhk7734 requested changes Feb 13, 2026

View reviewed changes

MAF-19283: docs(workflow): Update e2e-inference-perf-p-cluster.yaml t…

b35ae98

…o include result retrieval instructions for S3, enhancing clarity on post-test result management.

MAF-19283: refactor(e2e): Remove Heimdall tag references from environ…

ab61b26

…ment variables and test configurations to streamline setup and enhance clarity in performance and quality tests.

Merge branch 'main' into MAF-19283_add_toleration_affinity

d801a05

Copilot AI review requested due to automatic review settings February 13, 2026 01:31

Copilot started reviewing on behalf of hyeongyun0916 February 13, 2026 01:32 View session

Copilot AI reviewed Feb 13, 2026

View reviewed changes

seongsu-dev added 2 commits February 13, 2026 10:35

MAF-19283: refactor(e2e): Simplify inference performance and quality …

d49925e

…tests by introducing a conditional check for the IsKind variable, improving clarity in resource management and template rendering based on the environment configuration.

MAF-19283: docs(AGENTS): Clarify E2E test version scope and out-of-sc…

40e5bcb

…ope formats to improve testing guidelines

Copilot AI review requested due to automatic review settings February 13, 2026 01:38

Copilot started reviewing on behalf of ghost February 13, 2026 01:38 View session

Copilot AI reviewed Feb 13, 2026

View reviewed changes

hhk7734 approved these changes Feb 13, 2026

View reviewed changes

hhk7734 merged commit 0a44ee6 into main Feb 13, 2026
8 of 9 checks passed

hhk7734 deleted the MAF-19283_add_toleration_affinity branch February 13, 2026 06:10

Conversation

ghost commented Feb 12, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

ghost commented Feb 12, 2026 • edited by ghost Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

hyeongyun0916 commented Feb 13, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

hhk7734 left a comment

Choose a reason for hiding this comment

Uh oh!

ghost commented Feb 13, 2026

Uh oh!

ghost commented Feb 13, 2026

Uh oh!

ghost commented Feb 13, 2026

Uh oh!

hhk7734 commented Feb 13, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

ghost commented Feb 13, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

ghost commented Feb 12, 2026 •

edited by ghost

Loading

hyeongyun0916 commented Feb 13, 2026 •

edited

Loading