chore(release): retarget image and chart publish/pull to the llm-d org (llm-d#317)

The repo moved from llm-d-incubation to the llm-d org, but GHCR packages do not
follow a repo transfer, so every publish/pull path still targeted
ghcr.io/llm-d-incubation. Retarget them to ghcr.io/llm-d:

- ci-release.yaml / ci-build-images.yaml: build, push, and scan images under ghcr.io/llm-d
- scripts/publish-helm-chart.sh: push the chart to oci://ghcr.io/llm-d/charts
- charts/async-processor/values.yaml: default image pull repository -> ghcr.io/llm-d/llm-d-async
  (and the chart unit-test expectations to match)
- Makefile IMAGE_TAG_BASE, deploy/install.sh, the e2e default image, and the e2e-deploy docs

Go module paths (github.com/llm-d-incubation/...) are intentionally left unchanged:
they are self-consistent and resolve via GitHub's repo redirect; renaming them is a
separate breaking change for downstream consumers.

Verified: helm template renders ghcr.io/llm-d/llm-d-async, helm lint clean,
helm unittest 49/49 pass, go vet -tags e2e ok.

Signed-off-by: Shimi Bandiel <shimib@google.com>

Author: shimib

.github/workflows/ci-build-images.yaml

@@ -29,11 +29,11 @@ jobs:
           docker-file: Dockerfile
           tag: ${{ inputs.tag }}
           image-name: ${{ inputs.image-name }}
-          registry: ghcr.io/llm-d-incubation
+          registry: ghcr.io/llm-d
           github-token: ${{ secrets.GHCR_TOKEN }}
           prerelease: ${{ inputs.prerelease }}
 
       - name: Run Trivy scan on Async Processor image
         uses: ./.github/actions/trivy-scan
         with:
-          image: ghcr.io/llm-d-incubation/${{ inputs.image-name }}:${{ inputs.tag }}
+          image: ghcr.io/llm-d/${{ inputs.image-name }}:${{ inputs.tag }}

.github/workflows/ci-release.yaml

@@ -53,7 +53,7 @@ jobs:
 
       - name: Build and push multi-arch image
         run: |
-          IMAGE=ghcr.io/llm-d-incubation/${{ steps.version.outputs.project_name }}
+          IMAGE=ghcr.io/llm-d/${{ steps.version.outputs.project_name }}
           VERSION_TAG=${{ steps.tag.outputs.tag }}
           COMMIT_SHA=${{ steps.tag.outputs.commit_sha }}
           echo "Building $IMAGE with tags $VERSION_TAG and $COMMIT_SHA for linux/amd64,linux/arm64"
@@ -83,4 +83,4 @@ jobs:
       - name: Run Trivy scan
         uses: ./.github/actions/trivy-scan
         with:
-          image: ghcr.io/llm-d-incubation/${{ steps.version.outputs.project_name }}:${{ steps.tag.outputs.tag }}
+          image: ghcr.io/llm-d/${{ steps.version.outputs.project_name }}:${{ steps.tag.outputs.tag }}

Makefile

@@ -7,7 +7,7 @@ export
 
 
 # Image URL to use all building/pushing image targets
-IMAGE_TAG_BASE ?= ghcr.io/llm-d-incubation
+IMAGE_TAG_BASE ?= ghcr.io/llm-d
 IMG_TAG ?= latest
 IMG ?= $(IMAGE_TAG_BASE)/async-processor:$(IMG_TAG)
 
@@ -401,7 +401,7 @@ set-version:
 	done
 
 ## Copied from https://github.com/llm-d-incubation/batch-gateway
-## publish-helm-chart: Patch chart for VERSION, package, append chart to SHA256SUMS, push to oci://ghcr.io/llm-d-incubation/charts (requires VERSION, yq, helm; GITHUB_TOKEN, GITHUB_ACTOR for push).
+## publish-helm-chart: Patch chart for VERSION, package, append chart to SHA256SUMS, push to oci://ghcr.io/llm-d/charts (requires VERSION, yq, helm; GITHUB_TOKEN, GITHUB_ACTOR for push).
 .PHONY: publish-helm-chart
 publish-helm-chart:
 	@if [ -z "$(VERSION)" ]; then \

charts/async-processor/tests/deployment_test.yaml

@@ -163,15 +163,15 @@ tests:
     asserts:
       - equal:
           path: spec.template.spec.containers[0].image
-          value: ghcr.io/llm-d-incubation/llm-d-async:0.7.0
+          value: ghcr.io/llm-d/llm-d-async:0.7.0
 
   - it: should honor an explicit image tag override
     set:
       ap.image.tag: "v1.2.3"
     asserts:
       - equal:
           path: spec.template.spec.containers[0].image
-          value: ghcr.io/llm-d-incubation/llm-d-async:v1.2.3
+          value: ghcr.io/llm-d/llm-d-async:v1.2.3
 
   - it: should render --pubsub.project-id when projectId is set
     set:

charts/async-processor/values.yaml

@@ -20,7 +20,7 @@ ap:
       cpu: 500m
       memory: 512Mi
   image:
-    repository: ghcr.io/llm-d-incubation/llm-d-async
+    repository: ghcr.io/llm-d/llm-d-async
     # Image tag. Leave empty to track the chart's appVersion (recommended so the
     # image stays in sync with the chart). Set explicitly to pin a different tag.
     tag: ""

deploy/install.sh

@@ -31,7 +31,7 @@ REDIS_NS=${REDIS_NS:-"redis"}
 PROMETHEUS_SECRET_NS=${PROMETHEUS_SECRET_NS:-$MONITORING_NAMESPACE}
 
 # AP Configuration
-AP_IMAGE_REPO=${AP_IMAGE_REPO:-"ghcr.io/llm-d-incubation/async-processor"}
+AP_IMAGE_REPO=${AP_IMAGE_REPO:-"ghcr.io/llm-d/async-processor"}
 AP_IMAGE_TAG=${AP_IMAGE_TAG:-"latest"}
 AP_IMAGE_PULL_POLICY=${AP_IMAGE_PULL_POLICY:-"Never"}
 AP_RELEASE_NAME=${AP_RELEASE_NAME:-"async-processor"}

docs/guides/e2e-deploy.md

@@ -131,7 +131,7 @@ helm install async-processor ${ASYNC_REPO}/charts/async-processor/ \
 ```
 
 The values file (`docs/guides/e2e-deploy/async-processor-values.yaml`) configures:
-- Image: `ghcr.io/llm-d-incubation/llm-d-async:938cd44`
+- Image: `ghcr.io/llm-d/llm-d-async:938cd44`
 - Queue: Redis sorted-set with `redis.url` set directly (chart creates the Secret), configured via `queuesConfig`
 - Gate: `prometheus-budget` with pool=`optimized-baseline`, max_concurrency=100, baseline=0.05 (per-queue)
 - Prometheus URL pointing to the cluster's `llmd-kube-prometheus-stack-prometheus` service

docs/guides/e2e-deploy/async-processor-values.yaml

@@ -1,6 +1,6 @@
 ap:
   image:
-    repository: ghcr.io/llm-d-incubation/llm-d-async
+    repository: ghcr.io/llm-d/llm-d-async
     tag: "938cd44"
   imagePullPolicy: Always
   igwBaseURL: "http://llm-d-inference-gateway-istio:80"

scripts/publish-helm-chart.sh

@@ -7,7 +7,7 @@
 #   GITHUB_TOKEN  Token for helm registry login to ghcr.io
 #   GITHUB_ACTOR  Username for registry login (e.g. github.actor in Actions)
 #
-# Chart is always pushed to oci://ghcr.io/llm-d-incubation/charts (not configurable).
+# Chart is always pushed to oci://ghcr.io/llm-d/charts (not configurable).
 #
 # Requires: helm, yq (mikefarah). Run after make package-release so release/ exists.
 
@@ -21,7 +21,7 @@ CHART_VERSION="${VERSION#v}"
 export VERSION
 export CHART_VERSION
 
-HELM_OCI_REGISTRY='oci://ghcr.io/llm-d-incubation/charts'
+HELM_OCI_REGISTRY='oci://ghcr.io/llm-d/charts'
 
 SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
 REPO_ROOT="$(cd "${SCRIPT_DIR}/.." && pwd)"

test/e2e/e2e_suite_test.go

@@ -56,7 +56,7 @@ var (
 	jaegerPort     string = env.GetEnvString("E2E_INTEGRATION_JAEGER_PORT", "30494", ginkgo.GinkgoLogr)
 
 	containerRuntime = detectContainerRuntime()
-	apImage          = env.GetEnvString("AP_IMAGE", "ghcr.io/llm-d-incubation/async-processor:e2e-test", ginkgo.GinkgoLogr)
+	apImage          = env.GetEnvString("AP_IMAGE", "ghcr.io/llm-d/async-processor:e2e-test", ginkgo.GinkgoLogr)
 	eppImage         = env.GetEnvString("EPP_IMAGE", "registry.k8s.io/gateway-api-inference-extension/epp:v1.5.0", ginkgo.GinkgoLogr)
 	simImage         = env.GetEnvString("SIM_IMAGE", "ghcr.io/llm-d/llm-d-inference-sim:v0.9.1", ginkgo.GinkgoLogr)
 	gaieRoot         = os.Getenv("GAIE_ROOT")