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

Autoscaling Workflow Enhancement - Part 4: Integrating MetricClient into Autoscaling Workflow #116

Merged
merged 9 commits into from
Sep 6, 2024

Conversation

kr11
Copy link
Collaborator

@kr11 kr11 commented Sep 2, 2024

Pull Request Description

Compared to Part 3 (#101), this update introduces the metricsClient phase and completes the entire workflow, except the periodic resync AutoScaler invoking.

  1. Adding Prometheus Metrics Interface for Mocked Llama App:

    • Added a Prometheus metrics interface and test cases for the mocked llama-7b application located in docs/development/app.
    • The mocked llama-7b now first checks the latest number of replicas before returning metrics calculated as fix_total/replicas, thereby inversely relating metrics to the number of replicas. This change ensures that the scaling metrics dynamically adjust based on the current load distribution across replicas. See related issue [Enhance Mocked vLLM App with Dynamic Metrics for Autoscaling Efficiency #117] for more details.
  2. Integration of MetricClient into the Autoscaling Process:

    • Introduced a metricsClient into the autoscaling workflow, instantiated during the creation of the Reconciler.
    • The reconcile function now updates metrics before invoking the KPA algorithm. The KPA algorithm utilizes both current and historical metrics over a specified window to determine the appropriate scaling actions.
  3. Add tests and README.

Related Issues

Resolves: #117
TODO: #119

Important: Before submitting, please complete the description above and review the checklist below.


Contribution Guidelines (Expand for Details)

We appreciate your contribution to aibrix! To ensure a smooth review process and maintain high code quality, please adhere to the following guidelines:

Pull Request Title Format

Your PR title should start with one of these prefixes to indicate the nature of the change:

  • [Bug]: Corrections to existing functionality
  • [CI]: Changes to build process or CI pipeline
  • [Docs]: Updates or additions to documentation
  • [API]: Modifications to aibrix's API or interface
  • [CLI]: Changes or additions to the Command Line Interface
  • [Misc]: For changes not covered above (use sparingly)

Note: For changes spanning multiple categories, use multiple prefixes in order of importance.

Submission Checklist

  • PR title includes appropriate prefix(es)
  • Changes are clearly explained in the PR description
  • New and existing tests pass successfully
  • Code adheres to project style and best practices
  • Documentation updated to reflect changes (if applicable)
  • Thorough testing completed, no regressions introduced

By submitting this PR, you confirm that you've read these guidelines and your changes align with the project's contribution standards.

@kr11 kr11 force-pushed the kangrong/features/autoscaling_workflow_part4 branch from b487853 to 901bc86 Compare September 4, 2024 06:10
@kr11 kr11 changed the title WIP: Integrating MetricClient Module into Autoscaling workflow Autoscaling Workflow Enhancement - Part 4: Integrating MetricClient into Autoscaling Workflow Sep 4, 2024
@kr11 kr11 force-pushed the kangrong/features/autoscaling_workflow_part4 branch from 901bc86 to 3a96e20 Compare September 4, 2024 06:15
@kr11 kr11 force-pushed the kangrong/features/autoscaling_workflow_part4 branch from e28450b to ac50366 Compare September 6, 2024 12:34
@kr11
Copy link
Collaborator Author

kr11 commented Sep 6, 2024

I have fixed the conflict and rebased on up-to-date main branch. @Jeffwan

@Jeffwan
Copy link
Collaborator

Jeffwan commented Sep 6, 2024

the PR looks good to me

@Jeffwan Jeffwan merged commit be4900e into main Sep 6, 2024
3 checks passed
@Jeffwan Jeffwan deleted the kangrong/features/autoscaling_workflow_part4 branch September 6, 2024 13:37
gangmuk pushed a commit that referenced this pull request Jan 25, 2025
…nto Autoscaling Workflow (#116)

* enrich mocked vllm app

* update README

* Integrate metricsClient into KPA, Debug mode successful.

- Minor refactoring of Window and AutoScaler interfaces
- Adjust Reconcile workflow
- Add tests.

* add readme

* fix make lint-all

* fix readme, add more description

* add missing yaml

* optimize role name

* fix readme
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Enhance Mocked vLLM App with Dynamic Metrics for Autoscaling Efficiency
2 participants