✨ Document and test client-only install pattern (controller.enabled=false) by clubanderson · Pull Request #747 · llm-d/llm-d-autoscaling

clubanderson · 2026-02-17T00:18:38Z

Summary

Document the controller.enabled=false workflow for adding models to an existing WVA controller
Add Helm template tests verifying client-only vs full install resource generation

Fixes #746 (asks 2 and 3). Ask 1 (ConfigMap collision fix) is in #745.

Documentation (ask #2)

Adds a new "Adding Models to an Existing Controller" section to the multi-controller isolation guide with:

Step-by-step workflow: install controller once, add models with controller.enabled=false
Explanation of which resources are included/excluded in client-only mode
Tip about controllerInstance alignment between controller and model installs

Tests (ask #3)

Adds test/chart/client_only_install_test.go with 4 test cases using helm template:

Test	What it verifies
`TestClientOnlyInstall`	`controller.enabled=false` produces VA, HPA, Service, ServiceMonitor but excludes controller Deployment, ServiceAccount, leader election RBAC, ConfigMaps
`TestFullInstall`	`controller.enabled=true` includes all controller infrastructure
`TestClientOnlyNoVA`	Minimal install with no VA/HPA produces only Service/ServiceMonitor/RBAC
`TestClientOnlyControllerInstance`	`controllerInstance` label propagates to VA labels and HPA metric selectors in client-only mode

Run locally:

cd test/chart && go test -v ./...

Test plan

All 4 Helm template tests pass locally
Verify documentation renders correctly
CI checks pass

github-actions · 2026-02-17T00:21:21Z

GPU Pre-flight Check ✅

GPUs are available for e2e-openshift tests. Proceeding with deployment.

Resource	Total	Allocated	Available
GPUs	50	27	23

Cluster	Value
Nodes	16 (7 with GPUs)
Total CPU	993 cores
Total Memory	10383 Gi
GPUs required	4 (min) / 6 (recommended)

Copilot

Pull request overview

This PR documents and tests the controller.enabled=false workflow for adding models to an existing WVA controller without reinstalling the full stack. It addresses asks 2 and 3 from issue #746 by providing clear documentation and comprehensive Helm template tests.

Changes:

Adds "Adding Models to an Existing Controller" section to the multi-controller isolation guide with step-by-step instructions and resource breakdowns
Introduces 4 Helm template tests validating client-only install behavior, full install resources, minimal installs, and controllerInstance label propagation

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 2 comments.

File	Description
docs/user-guide/multi-controller-isolation.md	Documents controller-only install pattern with three-step workflow (install controller, add model A, add model B) and lists resources included/excluded in each mode
test/chart/client_only_install_test.go	Adds 4 test cases using `helm template` to verify resource generation for client-only vs full installs and controllerInstance label behavior

Copilot · 2026-02-17T00:23:34Z

+  --create-namespace \
+  --set controller.enabled=true \
+  --set va.enabled=false \
+  --set hpa.enabled=false


The controller-only install example should explicitly set vllmService.enabled=false to avoid creating unnecessary vLLM Service and ServiceMonitor resources with default model names. Since this step intends to install only the controller infrastructure without any workload-specific resources, the vLLM monitoring resources should also be disabled.

Suggested change

--set hpa.enabled=false

--set hpa.enabled=false \

--set vllmService.enabled=false

Copilot · 2026-02-17T00:23:34Z

@@ -0,0 +1,146 @@
+package chart_test


Missing Apache 2.0 license header. All Go files in this repository include a copyright and license header. Add the standard Apache 2.0 license header with "Copyright 2025." at the top of the file, consistent with other test files in the codebase.

bad AI. It's not missing, and you missed the fact that we are in 2026

github-actions · 2026-02-17T00:29:57Z

GPU Pre-flight Check ✅

GPUs are available for e2e-openshift tests. Proceeding with deployment.

Resource	Total	Allocated	Available
GPUs	119	101	18

Cluster	Value
Nodes	21 (15 with GPUs)
Total CPU	3391 cores
Total Memory	30350 Gi
GPUs required	4 (min) / 6 (recommended)

github-actions · 2026-02-17T01:22:46Z

GPU Pre-flight Check ✅

GPUs are available for e2e-openshift tests. Proceeding with deployment.

Resource	Total	Allocated	Available
GPUs	50	27	23

Cluster	Value
Nodes	16 (7 with GPUs)
Total CPU	993 cores
Total Memory	10383 Gi
GPUs required	4 (min) / 6 (recommended)

Copilot

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated 1 comment.

Copilot · 2026-02-17T01:35:54Z

+		{"leader-election", "leader election RBAC should be excluded"},
+		{"controller-manager", "controller manager resources should be excluded"},
+		{"prometheus-ca", "prometheus CA configmaps should be excluded"},
+		{"wva-deployment", "controller deployment template should be excluded"},


The marker "wva-deployment" doesn't appear anywhere in the Helm templates and therefore isn't a useful check. The deployment is already caught by the "controller-manager" marker on line 82, which appears in the deployment name, labels, and associated resources. Consider removing this redundant check or replacing it with a more specific marker if there's a particular resource you want to verify is excluded.

Suggested change

{"wva-deployment", "controller deployment template should be excluded"},

clubanderson · 2026-02-17T02:10:37Z

👋 @asm582 — all CI checks green. Could you approve when you get a chance? Thanks!

asm582

/lgtm

github-actions · 2026-02-17T02:26:13Z

GPU Pre-flight Check ✅

GPUs are available for e2e-openshift tests. Proceeding with deployment.

Resource	Total	Allocated	Available
GPUs	50	27	23

Cluster	Value
Nodes	16 (7 with GPUs)
Total CPU	993 cores
Total Memory	10383 Gi
GPUs required	4 (min) / 6 (recommended)

clubanderson · 2026-02-17T03:15:35Z

👋 @asm582 — Copilot comments addressed, all CI checks green (3/3 on e2e-openshift). Ready for your approval when you get a chance. Thanks!

lionelvillard · 2026-02-17T13:02:09Z

/hold

Having only one helm chart has been a great source of confusion for end-users. Let do this first: #555

@asm582

lionelvillard · 2026-02-17T20:38:10Z

/unhold

The merge-base changed after approval.

github-actions · 2026-02-17T21:48:03Z

GPU Pre-flight Check ✅

GPUs are available for e2e-openshift tests. Proceeding with deployment.

Resource	Total	Allocated	Available
GPUs	119	97	22

Cluster	Value
Nodes	21 (15 with GPUs)
Total CPU	3391 cores
Total Memory	30350 Gi
GPUs required	4 (min) / 6 (recommended)

lionelvillard · 2026-02-17T22:11:53Z

    fi
  fi
-
+  


please remove extra blank space, everywhere in this file

…alse) Add multi-controller isolation docs and Helm template tests for client-only install workflow. Signed-off-by: Andy Anderson <andy@clubanderson.com> Signed-off-by: Andrew Anderson <andy@clubanderson.com>

github-actions · 2026-02-17T22:18:58Z

GPU Pre-flight Check ❌

Insufficient GPUs to run e2e-openshift tests. Re-run with /retest when GPUs free up.

Resource	Total	Allocated	Available
GPUs	119	119	0

Cluster	Value
Nodes	21 (15 with GPUs)
Total CPU	3391 cores
Total Memory	30350 Gi
GPUs required	4 (min) / 6 (recommended)

Copilot

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated no new comments.

github-actions · 2026-02-17T22:33:37Z

GPU Pre-flight Check ❌

Insufficient GPUs to run e2e-openshift tests. Re-run with /retest when GPUs free up.

Resource	Total	Allocated	Available
GPUs	119	117	2

Cluster	Value
Nodes	21 (15 with GPUs)
Total CPU	3391 cores
Total Memory	30350 Gi
GPUs required	4 (min) / 6 (recommended)

github-actions · 2026-02-17T22:40:53Z

GPU Pre-flight Check ❌

Insufficient GPUs to run e2e-openshift tests. Re-run with /retest when GPUs free up.

Resource	Total	Allocated	Available
GPUs	119	120	-1

Cluster	Value
Nodes	21 (15 with GPUs)
Total CPU	3391 cores
Total Memory	30350 Gi
GPUs required	4 (min) / 6 (recommended)

github-actions · 2026-02-17T22:47:24Z

GPU Pre-flight Check ❌

Insufficient GPUs to run e2e-openshift tests. Re-run with /retest when GPUs free up.

Resource	Total	Allocated	Available
GPUs	119	117	2

Cluster	Value
Nodes	21 (15 with GPUs)
Total CPU	3391 cores
Total Memory	30350 Gi
GPUs required	4 (min) / 6 (recommended)

github-actions · 2026-02-17T22:50:19Z

GPU Pre-flight Check ❌

Insufficient GPUs to run e2e-openshift tests. Re-run with /retest when GPUs free up.

Resource	Total	Allocated	Available
GPUs	119	123	-4

Cluster	Value
Nodes	21 (15 with GPUs)
Total CPU	3391 cores
Total Memory	30350 Gi
GPUs required	4 (min) / 6 (recommended)

github-actions · 2026-02-17T22:58:41Z

GPU Pre-flight Check ✅

GPUs are available for e2e-openshift tests. Proceeding with deployment.

Resource	Total	Allocated	Available
GPUs	119	107	12

Cluster	Value
Nodes	21 (15 with GPUs)
Total CPU	3391 cores
Total Memory	30350 Gi
GPUs required	4 (min) / 6 (recommended)

github-actions · 2026-02-17T23:32:35Z

GPU Pre-flight Check ✅

GPUs are available for e2e-openshift tests. Proceeding with deployment.

Resource	Total	Allocated	Available
GPUs	119	111	8

Cluster	Value
Nodes	21 (15 with GPUs)
Total CPU	3391 cores
Total Memory	30350 Gi
GPUs required	4 (min) / 6 (recommended)

github-actions · 2026-02-18T00:02:46Z

GPU Pre-flight Check ✅

GPUs are available for e2e-openshift tests. Proceeding with deployment.

Resource	Total	Allocated	Available
GPUs	50	11	39

Cluster	Value
Nodes	16 (7 with GPUs)
Total CPU	993 cores
Total Memory	10383 Gi
GPUs required	4 (min) / 6 (recommended)

…alse) (llm-d#747) Add multi-controller isolation docs and Helm template tests for client-only install workflow. Signed-off-by: Andy Anderson <andy@clubanderson.com> Signed-off-by: Andrew Anderson <andy@clubanderson.com>

Copilot AI review requested due to automatic review settings February 17, 2026 00:18

Copilot started reviewing on behalf of clubanderson February 17, 2026 00:19 View session

Copilot AI reviewed Feb 17, 2026

View reviewed changes

Copilot AI review requested due to automatic review settings February 17, 2026 01:19

Copilot started reviewing on behalf of clubanderson February 17, 2026 01:20 View session

Copilot AI reviewed Feb 17, 2026

View reviewed changes

asm582 previously approved these changes Feb 17, 2026

View reviewed changes

clubanderson dismissed asm582’s stale review via 4187af5 February 17, 2026 02:23

clubanderson enabled auto-merge (squash) February 17, 2026 02:31

github-actions Bot added the hold PRs that are blocked on design, other features, release cycle, etc. label Feb 17, 2026

lionelvillard previously approved these changes Feb 17, 2026

View reviewed changes

clubanderson force-pushed the feat/client-only-install-docs-and-tests branch from 4187af5 to 4a92361 Compare February 17, 2026 21:45

lionelvillard reviewed Feb 17, 2026

View reviewed changes

Comment thread deploy/install.sh Outdated

fi

fi

lionelvillard Feb 17, 2026

Copy link
Copy Markdown

Collaborator

Choose a reason for hiding this comment

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

please remove extra blank space, everywhere in this file

✨ Document and test client-only install pattern (controller.enabled=f…

6ad2cb5

…alse) Add multi-controller isolation docs and Helm template tests for client-only install workflow. Signed-off-by: Andy Anderson <andy@clubanderson.com> Signed-off-by: Andrew Anderson <andy@clubanderson.com>

Copilot AI review requested due to automatic review settings February 17, 2026 22:16

clubanderson force-pushed the feat/client-only-install-docs-and-tests branch from 4a92361 to 6ad2cb5 Compare February 17, 2026 22:16

Copilot started reviewing on behalf of clubanderson February 17, 2026 22:16 View session

Copilot AI reviewed Feb 17, 2026

View reviewed changes

clubanderson closed this Feb 17, 2026

auto-merge was automatically disabled February 17, 2026 22:30
Pull request was closed

clubanderson reopened this Feb 17, 2026

github-actions Bot mentioned this pull request Feb 17, 2026

[agentics] No-Op Runs #744

Closed

lionelvillard removed the hold PRs that are blocked on design, other features, release cycle, etc. label Feb 18, 2026

lionelvillard approved these changes Feb 18, 2026

View reviewed changes

clubanderson merged commit 6563a7f into main Feb 18, 2026
67 of 74 checks passed

clubanderson deleted the feat/client-only-install-docs-and-tests branch February 18, 2026 12:43

clubanderson mentioned this pull request Feb 26, 2026

Governance, CI, and nightly E2E hygiene tracker (Feb 12–26) llm-d/llm-d#853

Closed

24 tasks

	--set hpa.enabled=false
	--set hpa.enabled=false \
	--set vllmService.enabled=false

Uh oh!

Conversation

clubanderson commented Feb 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Documentation (ask #2)

Tests (ask #3)

Test plan

Uh oh!

github-actions Bot commented Feb 17, 2026

GPU Pre-flight Check ✅

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Feb 17, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 17, 2026

Choose a reason for hiding this comment

Uh oh!

lionelvillard Feb 18, 2026

Choose a reason for hiding this comment

Uh oh!

github-actions Bot commented Feb 17, 2026

GPU Pre-flight Check ✅

Uh oh!

github-actions Bot commented Feb 17, 2026

GPU Pre-flight Check ✅

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Feb 17, 2026

Choose a reason for hiding this comment

Uh oh!

clubanderson commented Feb 17, 2026

Uh oh!

asm582 left a comment

Choose a reason for hiding this comment

Uh oh!

github-actions Bot commented Feb 17, 2026

GPU Pre-flight Check ✅

Uh oh!

clubanderson commented Feb 17, 2026

Uh oh!

lionelvillard commented Feb 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

lionelvillard commented Feb 17, 2026

Uh oh!

github-actions Bot commented Feb 17, 2026

GPU Pre-flight Check ✅

Uh oh!

lionelvillard Feb 17, 2026

Choose a reason for hiding this comment

Uh oh!

github-actions Bot commented Feb 17, 2026

GPU Pre-flight Check ❌

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

github-actions Bot commented Feb 17, 2026

GPU Pre-flight Check ❌

Uh oh!

github-actions Bot commented Feb 17, 2026

GPU Pre-flight Check ❌

Uh oh!

github-actions Bot commented Feb 17, 2026

GPU Pre-flight Check ❌

Uh oh!

github-actions Bot commented Feb 17, 2026

GPU Pre-flight Check ❌

Uh oh!

github-actions Bot commented Feb 17, 2026

clubanderson commented Feb 17, 2026 •

edited

Loading

lionelvillard commented Feb 17, 2026 •

edited

Loading