Wave3 #8

smarunich · 2025-07-15T16:31:32Z

This pull request introduces a comprehensive model publishing workflow for external access, along with updates to authentication, infrastructure, and documentation. The changes include defining requirements, implementing the publishing workflow, enhancing authentication mechanisms, and removing deprecated configurations.

Model Publishing Workflow:

.kiro/specs/model-publishing-workflow/requirements.md: Added a detailed requirements document outlining user stories, acceptance criteria, and features for publishing inference models securely, including tenant-specific access controls, API key generation, and rate limiting.
.kiro/specs/model-publishing-workflow/tasks.md: Defined an extensive implementation plan with tasks for setting up publishing infrastructure, extending Kubernetes client operations, creating API endpoints, and adding monitoring and UI components.

Authentication Enhancements:

management/auth.go: Introduced EnhancedAuthMiddleware to validate both JWT tokens and API keys, added API key validation logic, and implemented metadata search across namespaces for tenant-specific access.
management/main.go: Updated service initialization to include PublishingService and integrated it with the HTTP server setup.

Infrastructure Updates:

configs/management/management.yaml.bak: Removed deprecated configurations for the management-service, including service account, RBAC roles, and deployment specifications.

Documentation Improvements:

docs/getting-started.md: Updated the repository URL in the cloning instructions and added minor formatting improvements to the setup script description. [1] [2]

Minor Code Changes:

management/admin.go: Added the log package import to improve logging capabilities.
configs/kserve/models/huggingface-t5.yaml: Removed the --backend=huggingface argument from the model configuration.

smarunich · 2025-07-15T17:14:48Z

This pull request introduces significant enhancements to the model publishing workflow, including a detailed requirements document, an implementation plan, and updates to the authentication system. It also removes outdated configurations and adds new dependencies for better functionality. Below is a summary of the most important changes grouped by theme.

Enhancements to Model Publishing Workflow:

Added a comprehensive requirements document (.kiro/specs/model-publishing-workflow/requirements.md) outlining user stories and acceptance criteria for publishing models, tenant access, rate limiting, lifecycle management, documentation, and monitoring.
Introduced a detailed implementation plan (.kiro/specs/model-publishing-workflow/tasks.md) with tasks for infrastructure setup, API development, monitoring, UI components, and testing.

Authentication System Updates:

Enhanced AuthService to include Kubernetes client integration and support for validating API keys alongside JWT tokens (management/auth.go). Added EnhancedAuthMiddleware for dual authentication and ValidateAPIKey for API key validation. [1] [2]
Updated main.go to initialize PublishingService and integrate it with the HTTP server (management/main.go).

Removal of Outdated Configurations:

Removed deprecated Kubernetes configurations for the management service (configs/management/management.yaml.bak).

Minor Changes:

Updated repository URL in the "Getting Started" guide (docs/getting-started.md).
Added logging dependency to management/admin.go.

These changes collectively improve the functionality, security, and scalability of the model publishing workflow while simplifying the codebase and documentation.

Added 'configmaps' to the list of resources and enabled the 'watch' verb for core resources in the management.yaml RBAC configuration. This allows for broader monitoring and management capabilities.

smarunich · 2025-07-15T18:27:03Z

This pull request introduces significant updates to the model publishing workflow, including new requirements documentation, implementation plans, and enhancements to authentication mechanisms. It also includes smaller changes to configuration files and documentation. Below is a summary of the most important changes:

Model Publishing Workflow Enhancements

Requirements Documentation:

Added a comprehensive requirements document detailing user stories, acceptance criteria, and system capabilities for the model publishing workflow. This includes tenant-specific access control, API key generation, gateway routing, rate limiting, and lifecycle management. (.kiro/specs/model-publishing-workflow/requirements.md)

Implementation Plan:

Outlined detailed tasks for implementing the model publishing workflow, including API endpoints, gateway configurations, rate limiting policies, and monitoring/auditing features. (.kiro/specs/model-publishing-workflow/tasks.md)

Authentication Enhancements

Introduced an EnhancedAuthMiddleware in AuthService to support both JWT token and API key authentication. Added methods for validating API keys and retrieving metadata from Kubernetes secrets. (management/auth.go)
Updated AuthService to include a k8sClient for API key management operations. (management/auth.go)

Configuration Updates

Added configmaps to the list of Kubernetes resources with permissions for get, list, and watch in the management service configuration. (configs/management/management.yaml)
Removed the backup configuration file management.yaml.bak, which contained outdated service account and RBAC definitions. (configs/management/management.yaml.bak)

Documentation Updates

Updated the repository clone URL in the "Getting Started" guide to reflect the correct organization. (docs/getting-started.md)
Added a missing section to clarify the installation steps for dependencies in the "Getting Started" guide. (docs/getting-started.md)

Minor Code Changes

Added a log package import for improved logging capabilities. (management/admin.go)
Removed the --backend=huggingface argument from the HuggingFace T5 model configuration. (configs/kserve/models/huggingface-t5.yaml)

Copilot

Pull Request Overview

This PR implements a full end-to-end model publishing workflow, enhances authentication with API keys, removes deprecated management configs, and updates documentation.

Adds detailed publishing specs, tasks, and UI components for guided and quick publish flows
Introduces PublishingService with routes for publish, unpublish, key rotation, and validation
Cleans up old Kubernetes manifests and updates management-utils, server registration, and documentation

Reviewed Changes

Copilot reviewed 31 out of 31 changed files in this pull request and generated 5 comments.

Show a summary per file

File	Description
scripts/retest.sh	New test script to redeploy and port-forward the management service
scripts/build-and-push-images.sh	Removed deprecated manifest update helper
management/utils.go	Simplified command execution and YAML conversion, switched to yaml.v2
management/ui/src/index.css	Added CSS for publishing UI components and workflow indicators
management/ui/src/contexts/ApiContext.js	Exposed model publishing and API key management endpoints
management/types.go	Defined new types for publish configs, published models, and docs
management/server.go	Registered publishing routes and initialized `PublishingService`
management/publishing.go	Core implementation of the publishing workflow

Comments suppressed due to low confidence (1)

management/publishing.go:44

New core publishing logic is substantial and currently lacks unit or integration tests. Consider adding tests covering error paths, successful publish, and rollback scenarios.

func (s *PublishingService) PublishModel(c *gin.Context) {

Copilot · 2025-07-15T21:12:17Z

management/ui/src/contexts/ApiContext.js

+      const params = namespace ? { namespace } : {};
+      return api.post(`/models/${modelName}/publish/rotate-key`, {}, { params });
+    },
+    validateAPIKey: (apiKey) => api.post('/validate-api-key', { apiKey }),


The client is sending the API key in the request body but the server handler reads it from headers. Either adjust the client to set the X-API-Key header or update the server to parse the JSON body.

Suggested change

validateAPIKey: (apiKey) => api.post('/validate-api-key', { apiKey }),

validateAPIKey: (apiKey) => api.post('/validate-api-key', {}, { headers: { 'X-API-Key': apiKey } }),

Copilot · 2025-07-15T21:12:17Z

management/publishing.go

+								},
+							},
+							"limit": map[string]interface{}{
+								"requests": rateLimiting.RequestsPerMinute,


The rate-limiting policy only enforces requests per minute; the RequestsPerHour and BurstLimit fields are not applied. You should add additional rules for per-hour limits and burst handling.

Copilot · 2025-07-15T21:12:17Z

management/publishing.go

+	if rateLimiting.TokensPerHour > 0 {
+		rules := policy["spec"].(map[string]interface{})["rateLimit"].(map[string]interface{})["global"].(map[string]interface{})["rules"].([]interface{})
+
+		// Add token-based rate limiting


This section handles token limits but doesn’t enforce the configured RequestsPerHour or BurstLimit. Consider appending an additional rule for those values.

Suggested change

if rateLimiting.TokensPerHour > 0 {

rules := policy["spec"].(map[string]interface{})["rateLimit"].(map[string]interface{})["global"].(map[string]interface{})["rules"].([]interface{})

// Add token-based rate limiting

rules := policy["spec"].(map[string]interface{})["rateLimit"].(map[string]interface{})["global"].(map[string]interface{})["rules"].([]interface{})

// Add token-based rate limiting

if rateLimiting.TokensPerHour > 0 {

Copilot · 2025-07-15T21:12:18Z

management/ui/src/index.css

+  color: #2563eb;
+}
+
+.form-section {


[nitpick] The .form-section selector is defined twice (around line 101 and here). Consolidate duplicate rules to reduce CSS bloat and avoid conflicting styles.

Copilot · 2025-07-15T21:12:18Z

management/ui/src/components/PublishingWorkflow.js

+  };
+
+  const renderStepIndicator = () => (
+    <div style={{ 


[nitpick] There are large inline style blocks for the step indicator; consider moving these into CSS classes to improve readability and reuse.

smarunich and others added 5 commits July 15, 2025 09:16

Update huggingface-t5.yaml

0c6fe13

fix logger

9c24eef

fix admin resources

e97619e

javascript empty outputs handling

6d537cb

adding publishing workflow

940d14d

smarunich requested a review from Copilot July 15, 2025 16:49

This comment was marked as outdated.

Sign in to view

smarunich added 2 commits July 15, 2025 16:56

UI

8c3a561

PR comments

317b4a6

smarunich requested a review from Copilot July 15, 2025 17:05

This comment was marked as outdated.

Sign in to view

smarunich added 2 commits July 15, 2025 17:10

typo

fe22cbc

minor fixes

bcfe28f

smarunich requested a review from Copilot July 15, 2025 17:14

This comment was marked as outdated.

Sign in to view

smarunich and others added 2 commits July 15, 2025 13:21

Update RBAC rules for additional resource access

e964be4

Added 'configmaps' to the list of resources and enabled the 'watch' verb for core resources in the management.yaml RBAC configuration. This allows for broader monitoring and management capabilities.

expanding the functionality

9a6e584

smarunich requested a review from Copilot July 15, 2025 18:26

This comment was marked as outdated.

Sign in to view

smarunich added 3 commits July 15, 2025 18:46

add documentation

b8864ef

typo

49c8b47

typos

50043ec

smarunich requested a review from Copilot July 15, 2025 18:54

more typos

085d9ae

This comment was marked as outdated.

Sign in to view

smarunich added 3 commits July 15, 2025 18:56

fixes

b1f2ff8

more fixes

e175ac9

expose dashboard

abf858e

smarunich added 3 commits July 15, 2025 19:24

expanding publishing workflow

6d986f0

fix auth validation

43a8674

updates

e46afe9

smarunich requested a review from Copilot July 15, 2025 20:00

publishing updates

63b5ad6

This comment was marked as outdated.

Sign in to view

smarunich added 7 commits July 15, 2025 20:13

typo fixes

a2c42fc

fix bug

9e9036a

more updates

c5b0407

more updates to publsihing

42d69c2

delete updates

1f7ace7

UI concerns

8e90331

retest.sh

5b81f10

smarunich requested a review from Copilot July 15, 2025 21:09

fix typo

09663fc

Copilot AI reviewed Jul 15, 2025

View reviewed changes

smarunich merged commit 7860b29 into main Jul 15, 2025
2 of 5 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Wave3 #8

Wave3 #8

Uh oh!

smarunich commented Jul 15, 2025 •

edited

Loading

Uh oh!

This comment was marked as outdated.

Uh oh!

This comment was marked as outdated.

Uh oh!

smarunich commented Jul 15, 2025

Uh oh!

This comment was marked as outdated.

Uh oh!

smarunich commented Jul 15, 2025

Uh oh!

This comment was marked as outdated.

Uh oh!

This comment was marked as outdated.

Uh oh!

This comment was marked as outdated.

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Jul 15, 2025

Uh oh!

Copilot AI Jul 15, 2025

Uh oh!

Copilot AI Jul 15, 2025

Uh oh!

Copilot AI Jul 15, 2025

Uh oh!

Copilot AI Jul 15, 2025

Uh oh!

Uh oh!

Uh oh!

	validateAPIKey: (apiKey) => api.post('/validate-api-key', { apiKey }),
	validateAPIKey: (apiKey) => api.post('/validate-api-key', {}, { headers: { 'X-API-Key': apiKey } }),

Wave3 #8

Wave3 #8

Uh oh!

Conversation

smarunich commented Jul 15, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Model Publishing Workflow:

Authentication Enhancements:

Infrastructure Updates:

Documentation Improvements:

Minor Code Changes:

Uh oh!

This comment was marked as outdated.

Uh oh!

This comment was marked as outdated.

Uh oh!

smarunich commented Jul 15, 2025

Enhancements to Model Publishing Workflow:

Authentication System Updates:

Removal of Outdated Configurations:

Minor Changes:

Uh oh!

This comment was marked as outdated.

Uh oh!

smarunich commented Jul 15, 2025

Model Publishing Workflow Enhancements

Requirements Documentation:

Implementation Plan:

Authentication Enhancements

Configuration Updates

Documentation Updates

Minor Code Changes

Uh oh!

This comment was marked as outdated.

Uh oh!

This comment was marked as outdated.

Uh oh!

This comment was marked as outdated.

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Reviewed Changes

Uh oh!

Copilot AI Jul 15, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jul 15, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jul 15, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jul 15, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jul 15, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

smarunich commented Jul 15, 2025 •

edited

Loading