diff --git a/.github/workflows/container-validation-backends.yml b/.github/workflows/container-validation-backends.yml index f7f0a9703128..0d250eebba79 100644 --- a/.github/workflows/container-validation-backends.yml +++ b/.github/workflows/container-validation-backends.yml @@ -160,17 +160,14 @@ jobs: sglang: needs: changed-files if: needs.changed-files.outputs.has_code_changes == 'true' - # OPS-1140: Uncomment this for sglang arm switch to wideep - # strategy: - # fail-fast: false - # matrix: - # platform: - # - { arch: amd64, runner: gpu-l40-amd64 } - # - { arch: arm64, runner: cpu-arm-r8g-4xlarge } - # name: sglang (${{ matrix.platform.arch }}) - # runs-on: ${{ matrix.platform.runner }} - # OPS-1140: Remove this runs-on line, replaced with the above line - runs-on: gpu-l40-amd64 + strategy: + fail-fast: false + matrix: + platform: + - { arch: amd64, runner: gpu-l40-amd64 } + - { arch: arm64, runner: cpu-arm-r8g-4xlarge } + name: sglang (${{ matrix.platform.arch }}) + runs-on: ${{ matrix.platform.runner }} steps: - name: Checkout repository uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # v4.3.0 @@ -181,9 +178,7 @@ jobs: with: framework: sglang target: runtime - platform: 'linux/amd64' - # OPS-1140: Replace the above line with the uncommented below line - # platform: 'linux/${{ matrix.platform.arch }}' + platform: 'linux/${{ matrix.platform.arch }}' ngc_ci_access_token: ${{ secrets.NGC_CI_ACCESS_TOKEN }} ci_token: ${{ secrets.CI_TOKEN }} aws_default_region: ${{ secrets.AWS_DEFAULT_REGION }} @@ -196,9 +191,7 @@ jobs: uses: ./.github/actions/docker-tag-push with: local_image: ${{ steps.build-image.outputs.image_tag }} - push_tag: ai-dynamo/dynamo:${{ github.sha }}-sglang-amd64 - # OPS-1140: Replace the above line with the uncommented below line - # push_tag: ai-dynamo/dynamo:${{ github.sha }}-sglang-${{ matrix.platform.arch }} + push_tag: ai-dynamo/dynamo:${{ github.sha }}-sglang-${{ matrix.platform.arch }} # OPS-1145: Switch aws_push to true aws_push: 'false' azure_push: 'true' @@ -209,8 +202,7 @@ jobs: azure_acr_password: ${{ secrets.AZURE_ACR_PASSWORD }} - name: Run unit tests - # OPS-1140: Uncomment the below line - # if: ${{ matrix.platform.arch != 'arm64' }} + if: ${{ matrix.platform.arch != 'arm64' }} uses: ./.github/actions/pytest with: image_tag: ${{ steps.build-image.outputs.image_tag }} @@ -219,8 +211,7 @@ jobs: test_type: "unit" platform_arch: ${{ matrix.platform.arch }} - name: Run e2e tests - # OPS-1140: Uncomment the below line - # if: ${{ matrix.platform.arch != 'arm64' }} + if: ${{ matrix.platform.arch != 'arm64' }} uses: ./.github/actions/pytest with: image_tag: ${{ steps.build-image.outputs.image_tag }} diff --git a/Earthfile b/Earthfile index e59798d87aa6..57a307cbe69e 100644 --- a/Earthfile +++ b/Earthfile @@ -134,7 +134,7 @@ dynamo-build: dynamo-base-docker: ARG IMAGE=dynamo-base-docker - ARG DOCKER_SERVER=my-registry + ARG DOCKER_SERVER=nvcr.io/nvidia/ai-dynamo ARG IMAGE_TAG=latest FROM ubuntu:24.04 @@ -175,7 +175,7 @@ all-test: BUILD ./deploy/cloud/operator+test all-docker: - ARG DOCKER_SERVER=my-registry + ARG DOCKER_SERVER=nvcr.io/nvidia/ai-dynamo ARG IMAGE_TAG=latest BUILD ./deploy/cloud/operator+docker --DOCKER_SERVER=$DOCKER_SERVER --IMAGE_TAG=$IMAGE_TAG @@ -189,6 +189,6 @@ all: # For testing custom: - ARG DOCKER_SERVER=my-registry + ARG DOCKER_SERVER=nvcr.io/nvidia/ai-dynamo ARG IMAGE_TAG=latest BUILD +all-test diff --git a/benchmarks/incluster/benchmark_job.yaml b/benchmarks/incluster/benchmark_job.yaml index b81a65895a81..a59a4523af6b 100644 --- a/benchmarks/incluster/benchmark_job.yaml +++ b/benchmarks/incluster/benchmark_job.yaml @@ -17,7 +17,7 @@ spec: containers: - name: benchmark-runner # TODO: update to latest public image in next release - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 securityContext: allowPrivilegeEscalation: false capabilities: diff --git a/benchmarks/profiler/utils/config.py b/benchmarks/profiler/utils/config.py index d04b11fab41f..9a2a2f1f2a32 100644 --- a/benchmarks/profiler/utils/config.py +++ b/benchmarks/profiler/utils/config.py @@ -107,7 +107,7 @@ class DgdPlannerServiceConfig(BaseModel): volumeMounts: list[VolumeMount] = [VolumeMount()] extraPodSpec: PodSpec = PodSpec( mainContainer=Container( - image="my-registry/dynamo-runtime:my-tag", # placeholder + image="nvcr.io/nvidia/ai-dynamo/dynamo-runtime:0.6.1", # placeholder workingDir=f"{get_workspace_dir()}/components/src/dynamo/planner", command=["python3", "-m", "planner_sla"], args=[], diff --git a/components/backends/sglang/deploy/README.md b/components/backends/sglang/deploy/README.md index 8b77cd973ebc..55bb94e4204c 100644 --- a/components/backends/sglang/deploy/README.md +++ b/components/backends/sglang/deploy/README.md @@ -61,7 +61,7 @@ resources: ```yaml extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 workingDir: /workspace/components/backends/sglang args: - "python3" @@ -92,7 +92,7 @@ Edit the template to match your environment: ```yaml # Update image registry and tag -image: my-registry/sglang-runtime:my-tag +image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 # Configure your model args: diff --git a/components/backends/sglang/deploy/agg.yaml b/components/backends/sglang/deploy/agg.yaml index 8c444b384e17..c9856a3d0abd 100644 --- a/components/backends/sglang/deploy/agg.yaml +++ b/components/backends/sglang/deploy/agg.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 decode: envFromSecret: hf-token-secret dynamoNamespace: sglang-agg @@ -24,7 +24,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 workingDir: /workspace/components/backends/sglang command: - python3 diff --git a/components/backends/sglang/deploy/agg_logging.yaml b/components/backends/sglang/deploy/agg_logging.yaml index 93fa747c2bef..eff2aabc3b6f 100644 --- a/components/backends/sglang/deploy/agg_logging.yaml +++ b/components/backends/sglang/deploy/agg_logging.yaml @@ -16,7 +16,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 decode: envFromSecret: hf-token-secret dynamoNamespace: sglang-agg @@ -27,7 +27,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 workingDir: /workspace/components/backends/sglang command: - python3 diff --git a/components/backends/sglang/deploy/agg_router.yaml b/components/backends/sglang/deploy/agg_router.yaml index 142f1932f82f..fe93a5d96fc0 100644 --- a/components/backends/sglang/deploy/agg_router.yaml +++ b/components/backends/sglang/deploy/agg_router.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 envs: - name: DYN_ROUTER_MODE value: kv @@ -27,7 +27,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 workingDir: /workspace/components/backends/sglang command: - python3 diff --git a/components/backends/sglang/deploy/disagg-multinode.yaml b/components/backends/sglang/deploy/disagg-multinode.yaml index 39806f7ba653..3a015f27d173 100644 --- a/components/backends/sglang/deploy/disagg-multinode.yaml +++ b/components/backends/sglang/deploy/disagg-multinode.yaml @@ -22,7 +22,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 decode: multinode: nodeCount: 2 @@ -35,7 +35,7 @@ spec: gpu: "4" extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 workingDir: /workspace/components/backends/sglang command: - python3 @@ -72,7 +72,7 @@ spec: gpu: "4" extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 workingDir: /workspace/components/backends/sglang command: - python3 diff --git a/components/backends/sglang/deploy/disagg.yaml b/components/backends/sglang/deploy/disagg.yaml index d162844f4c8c..7cee539bb259 100644 --- a/components/backends/sglang/deploy/disagg.yaml +++ b/components/backends/sglang/deploy/disagg.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 decode: envFromSecret: hf-token-secret dynamoNamespace: sglang-disagg @@ -25,7 +25,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 workingDir: /workspace/components/backends/sglang command: - python3 @@ -61,7 +61,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 workingDir: /workspace/components/backends/sglang command: - python3 diff --git a/components/backends/sglang/deploy/disagg_planner.yaml b/components/backends/sglang/deploy/disagg_planner.yaml index 5e0d7f2368bb..e0cd608ddfe3 100644 --- a/components/backends/sglang/deploy/disagg_planner.yaml +++ b/components/backends/sglang/deploy/disagg_planner.yaml @@ -16,7 +16,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 Planner: dynamoNamespace: dynamo envFromSecret: hf-token-secret @@ -27,7 +27,7 @@ spec: mountPoint: /data extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 workingDir: /workspace/components/src/dynamo/planner command: - python3 @@ -49,7 +49,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 workingDir: /workspace/components/backends/sglang command: - python3 @@ -85,7 +85,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 workingDir: /workspace/components/backends/sglang command: - python3 diff --git a/components/backends/trtllm/deploy/README.md b/components/backends/trtllm/deploy/README.md index 8fad703d9680..11769e724146 100644 --- a/components/backends/trtllm/deploy/README.md +++ b/components/backends/trtllm/deploy/README.md @@ -89,7 +89,7 @@ resources: ```yaml extraPodSpec: mainContainer: - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 workingDir: /workspace/components/backends/trtllm args: - "python3" @@ -109,7 +109,7 @@ Before using these templates, ensure you have: ### Container Images -The deployment files currently require access to `my-registry/trtllm-runtime`. If you don't have access, build and push your own image: +The deployment files currently require access to `nvcr.io/nvidia/ai-dynamo/trtllm-runtime`. If you don't have access, build and push your own image: ```bash ./container/build.sh --framework tensorrtllm @@ -141,7 +141,7 @@ Edit the template to match your environment: ```yaml # Update image registry and tag -image: my-registry/trtllm-runtime:my-tag +image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 # Configure your model and deployment settings args: diff --git a/components/backends/trtllm/deploy/agg-with-config.yaml b/components/backends/trtllm/deploy/agg-with-config.yaml index e40ca48ada33..e4c10a0fb93b 100644 --- a/components/backends/trtllm/deploy/agg-with-config.yaml +++ b/components/backends/trtllm/deploy/agg-with-config.yaml @@ -34,7 +34,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:0.6.1 TRTLLMWorker: envFromSecret: hf-token-secret dynamoNamespace: trtllm-agg @@ -50,7 +50,7 @@ spec: configMap: name: nvidia-config mainContainer: - image: nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:0.6.1 workingDir: /workspace/components/backends/trtllm # mount the configmap as a volume volumeMounts: diff --git a/components/backends/trtllm/deploy/agg.yaml b/components/backends/trtllm/deploy/agg.yaml index 54412576a258..682d0765d2e5 100644 --- a/components/backends/trtllm/deploy/agg.yaml +++ b/components/backends/trtllm/deploy/agg.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 TRTLLMWorker: envFromSecret: hf-token-secret dynamoNamespace: trtllm-agg @@ -24,7 +24,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 workingDir: /workspace/ command: - python3 diff --git a/components/backends/trtllm/deploy/agg_router.yaml b/components/backends/trtllm/deploy/agg_router.yaml index ed42129fb421..d1a27cc7ed01 100644 --- a/components/backends/trtllm/deploy/agg_router.yaml +++ b/components/backends/trtllm/deploy/agg_router.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 envs: - name: DYN_ROUTER_MODE value: kv @@ -27,7 +27,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 workingDir: /workspace/ command: - python3 diff --git a/components/backends/trtllm/deploy/disagg-multinode.yaml b/components/backends/trtllm/deploy/disagg-multinode.yaml index 2906cfd193e6..adf31e4f5a70 100644 --- a/components/backends/trtllm/deploy/disagg-multinode.yaml +++ b/components/backends/trtllm/deploy/disagg-multinode.yaml @@ -95,7 +95,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 workingDir: /workspace/components/backends/trtllm command: - python3 @@ -127,7 +127,7 @@ spec: - name: nvidia-config mountPath: /workspace/ readOnly: true - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 workingDir: /workspace/ command: - python3 @@ -167,7 +167,7 @@ spec: - name: nvidia-config mountPath: /workspace/ readOnly: true - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 workingDir: /workspace/ command: - python3 diff --git a/components/backends/trtllm/deploy/disagg.yaml b/components/backends/trtllm/deploy/disagg.yaml index 501d2a4c200e..fff9433bf220 100644 --- a/components/backends/trtllm/deploy/disagg.yaml +++ b/components/backends/trtllm/deploy/disagg.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 TRTLLMPrefillWorker: dynamoNamespace: trtllm-disagg envFromSecret: hf-token-secret @@ -25,7 +25,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 workingDir: /workspace/ command: - python3 @@ -53,7 +53,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 workingDir: /workspace/ command: - python3 diff --git a/components/backends/trtllm/deploy/disagg_planner.yaml b/components/backends/trtllm/deploy/disagg_planner.yaml index 40c5a82ecf3a..a59cf98cc2fa 100644 --- a/components/backends/trtllm/deploy/disagg_planner.yaml +++ b/components/backends/trtllm/deploy/disagg_planner.yaml @@ -16,7 +16,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 workingDir: /workspace/components/backends/trtllm command: - python3 @@ -44,7 +44,7 @@ spec: mountPoint: /data extraPodSpec: mainContainer: - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 workingDir: /workspace/components/src/dynamo/planner ports: - name: metrics @@ -85,7 +85,7 @@ spec: extraPodSpec: terminationGracePeriodSeconds: 600 mainContainer: - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 workingDir: /workspace/ command: - python3 @@ -114,7 +114,7 @@ spec: extraPodSpec: terminationGracePeriodSeconds: 600 mainContainer: - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 workingDir: /workspace/ command: - python3 diff --git a/components/backends/trtllm/deploy/disagg_router.yaml b/components/backends/trtllm/deploy/disagg_router.yaml index f687354a9546..7eb26cc1cd12 100644 --- a/components/backends/trtllm/deploy/disagg_router.yaml +++ b/components/backends/trtllm/deploy/disagg_router.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 envs: - name: DYN_ROUTER_MODE value: kv @@ -27,7 +27,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 workingDir: /workspace/ command: - python3 @@ -55,7 +55,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 workingDir: /workspace/ command: - python3 diff --git a/components/backends/vllm/deploy/README.md b/components/backends/vllm/deploy/README.md index 16169381d1a1..ca4769634d67 100644 --- a/components/backends/vllm/deploy/README.md +++ b/components/backends/vllm/deploy/README.md @@ -69,7 +69,7 @@ resources: ```yaml extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm args: - "python3" @@ -116,7 +116,7 @@ Edit the template to match your environment: ```yaml # Update image registry and tag -image: my-registry/vllm-runtime:my-tag +image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 # Configure your model args: diff --git a/components/backends/vllm/deploy/agg.yaml b/components/backends/vllm/deploy/agg.yaml index 95de87138a4e..67b7f1159ae4 100644 --- a/components/backends/vllm/deploy/agg.yaml +++ b/components/backends/vllm/deploy/agg.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 VllmDecodeWorker: envFromSecret: hf-token-secret dynamoNamespace: vllm-agg @@ -24,7 +24,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 diff --git a/components/backends/vllm/deploy/agg_kvbm.yaml b/components/backends/vllm/deploy/agg_kvbm.yaml index 84dc29778633..afffff90d950 100644 --- a/components/backends/vllm/deploy/agg_kvbm.yaml +++ b/components/backends/vllm/deploy/agg_kvbm.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 VllmDecodeWorker: envFromSecret: hf-token-secret dynamoNamespace: vllm-agg-kvbm @@ -31,7 +31,7 @@ spec: value: "100" extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 diff --git a/components/backends/vllm/deploy/agg_router.yaml b/components/backends/vllm/deploy/agg_router.yaml index 999dd75f6411..aa2197f5022d 100644 --- a/components/backends/vllm/deploy/agg_router.yaml +++ b/components/backends/vllm/deploy/agg_router.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 envs: - name: DYN_ROUTER_MODE value: kv @@ -27,7 +27,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 diff --git a/components/backends/vllm/deploy/disagg-multinode.yaml b/components/backends/vllm/deploy/disagg-multinode.yaml index e46b2ed6afc5..7ad9b72d3c15 100644 --- a/components/backends/vllm/deploy/disagg-multinode.yaml +++ b/components/backends/vllm/deploy/disagg-multinode.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 @@ -34,7 +34,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 @@ -57,7 +57,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 diff --git a/components/backends/vllm/deploy/disagg.yaml b/components/backends/vllm/deploy/disagg.yaml index d7288a62da0f..edad31af5578 100644 --- a/components/backends/vllm/deploy/disagg.yaml +++ b/components/backends/vllm/deploy/disagg.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 VllmDecodeWorker: dynamoNamespace: vllm-disagg envFromSecret: hf-token-secret @@ -25,7 +25,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 @@ -45,7 +45,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 diff --git a/components/backends/vllm/deploy/disagg_kvbm.yaml b/components/backends/vllm/deploy/disagg_kvbm.yaml index d3455a87ce8a..8b130ed3fb2d 100644 --- a/components/backends/vllm/deploy/disagg_kvbm.yaml +++ b/components/backends/vllm/deploy/disagg_kvbm.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 VllmDecodeWorker: dynamoNamespace: vllm-disagg-kvbm envFromSecret: hf-token-secret @@ -24,7 +24,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 @@ -56,7 +56,7 @@ spec: value: "100" extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 diff --git a/components/backends/vllm/deploy/disagg_kvbm_2p2d.yaml b/components/backends/vllm/deploy/disagg_kvbm_2p2d.yaml index 883a64b02bef..e6719b3d2cc2 100644 --- a/components/backends/vllm/deploy/disagg_kvbm_2p2d.yaml +++ b/components/backends/vllm/deploy/disagg_kvbm_2p2d.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 VllmDecodeWorker: dynamoNamespace: vllm-disagg-kvbm-2p2d envFromSecret: hf-token-secret @@ -24,7 +24,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 @@ -56,7 +56,7 @@ spec: value: "100" extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 diff --git a/components/backends/vllm/deploy/disagg_kvbm_tp2.yaml b/components/backends/vllm/deploy/disagg_kvbm_tp2.yaml index 82b9c59f932d..e5cdb99b51cc 100644 --- a/components/backends/vllm/deploy/disagg_kvbm_tp2.yaml +++ b/components/backends/vllm/deploy/disagg_kvbm_tp2.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 VllmDecodeWorker: dynamoNamespace: vllm-disagg-kvbm-tp2 envFromSecret: hf-token-secret @@ -26,7 +26,7 @@ spec: gpu: "2" extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 @@ -60,7 +60,7 @@ spec: value: "100" extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 diff --git a/components/backends/vllm/deploy/disagg_planner.yaml b/components/backends/vllm/deploy/disagg_planner.yaml index 5afbf58c70d9..6c2582ffc0da 100644 --- a/components/backends/vllm/deploy/disagg_planner.yaml +++ b/components/backends/vllm/deploy/disagg_planner.yaml @@ -16,7 +16,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 Planner: dynamoNamespace: vllm-disagg-planner componentType: planner @@ -26,7 +26,7 @@ spec: mountPoint: /data extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/src/dynamo/planner command: - python3 @@ -48,7 +48,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 @@ -68,7 +68,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 diff --git a/components/backends/vllm/deploy/disagg_router.yaml b/components/backends/vllm/deploy/disagg_router.yaml index a298bdcfe519..906d6308992a 100644 --- a/components/backends/vllm/deploy/disagg_router.yaml +++ b/components/backends/vllm/deploy/disagg_router.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 envs: - name: DYN_ROUTER_MODE value: kv @@ -27,7 +27,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 @@ -46,7 +46,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 diff --git a/deploy/cloud/helm/platform/Chart.yaml b/deploy/cloud/helm/platform/Chart.yaml index 49744a5a2215..41f392c7a756 100644 --- a/deploy/cloud/helm/platform/Chart.yaml +++ b/deploy/cloud/helm/platform/Chart.yaml @@ -19,11 +19,11 @@ maintainers: url: https://www.nvidia.com description: A Helm chart for NVIDIA Dynamo Platform. type: application -version: 0.6.0 +version: 0.6.1 home: https://nvidia.com dependencies: - name: dynamo-operator - version: 0.5.0 + version: 0.6.1 repository: file://components/operator condition: dynamo-operator.enabled - name: nats diff --git a/deploy/cloud/operator/Earthfile b/deploy/cloud/operator/Earthfile index 4d368e12b52a..c80fe353f06c 100644 --- a/deploy/cloud/operator/Earthfile +++ b/deploy/cloud/operator/Earthfile @@ -45,7 +45,7 @@ test: SAVE ARTIFACT cover.out docker: - ARG DOCKER_SERVER=my-registry + ARG DOCKER_SERVER=nvcr.io/nvidia/ai-dynamo ARG IMAGE_TAG=latest ARG IMAGE_SUFFIX=dynamo-operator FROM nvcr.io/nvidia/distroless/go:v3.1.13 diff --git a/deploy/cloud/operator/internal/secrets/docker_test.go b/deploy/cloud/operator/internal/secrets/docker_test.go index 64d05890e0d1..faa652682e8e 100644 --- a/deploy/cloud/operator/internal/secrets/docker_test.go +++ b/deploy/cloud/operator/internal/secrets/docker_test.go @@ -21,7 +21,7 @@ func TestDockerSecretIndexer_RefreshIndex(t *testing.T) { }, Type: corev1.SecretTypeDockerConfigJson, Data: map[string][]byte{ - ".dockerconfigjson": []byte(`{"auths":{"docker.io":{}, "my-registry.com:5005/registry1":{}}}`), + ".dockerconfigjson": []byte(`{"auths":{"docker.io":{}, "nvcr.io/nvidia/ai-dynamo.com:5005/registry1":{}}}`), }, }, { @@ -31,7 +31,7 @@ func TestDockerSecretIndexer_RefreshIndex(t *testing.T) { }, Type: corev1.SecretTypeDockerConfigJson, Data: map[string][]byte{ - ".dockerconfigjson": []byte(`{"auths":{"my-registry.com:5005/registry2":{}}}`), + ".dockerconfigjson": []byte(`{"auths":{"nvcr.io/nvidia/ai-dynamo.com:5005/registry2":{}}}`), }, }, { @@ -41,7 +41,7 @@ func TestDockerSecretIndexer_RefreshIndex(t *testing.T) { }, Type: corev1.SecretTypeDockerConfigJson, Data: map[string][]byte{ - ".dockerconfigjson": []byte(`{"auths":{"my-registry.com:5005/registry2":{}}}`), + ".dockerconfigjson": []byte(`{"auths":{"nvcr.io/nvidia/ai-dynamo.com:5005/registry2":{}}}`), }, }, } @@ -68,7 +68,7 @@ func TestDockerSecretIndexer_RefreshIndex(t *testing.T) { t.Errorf("DockerSecretIndexer.GetSecrets() = %v, want %v", secrets[0], "secret1") } - secrets, err = i.GetSecrets("default", "my-registry.com:5005") + secrets, err = i.GetSecrets("default", "nvcr.io/nvidia/ai-dynamo.com:5005") if err != nil { t.Errorf("DockerSecretIndexer.GetSecrets() error = %v, wantErr %v", err, nil) } @@ -82,7 +82,7 @@ func TestDockerSecretIndexer_RefreshIndex(t *testing.T) { t.Errorf("DockerSecretIndexer.GetSecrets() = %v, want %v", secrets[1], "secret2") } - secrets, err = i.GetSecrets("another-namespace", "my-registry.com:5005") + secrets, err = i.GetSecrets("another-namespace", "nvcr.io/nvidia/ai-dynamo.com:5005") if err != nil { t.Errorf("DockerSecretIndexer.GetSecrets() error = %v, wantErr %v", err, nil) } diff --git a/deploy/cloud/pre-deployment/nixl/README.md b/deploy/cloud/pre-deployment/nixl/README.md index 051624e646de..140efea36176 100644 --- a/deploy/cloud/pre-deployment/nixl/README.md +++ b/deploy/cloud/pre-deployment/nixl/README.md @@ -286,7 +286,7 @@ Interactive script that provides flexible build and deployment workflow: ### nixlbench-deployment.yaml Base Kubernetes deployment template that gets customized by the script: -- **Template image**: `my-registry/nixlbench:version-arch` +- **Template image**: `nvcr.io/nvidia/ai-dynamo/nixlbench:version-arch` - **Resource allocation**: 10 CPU, 5Gi memory, 1 GPU per pod - **ETCD integration**: Pre-configured environment variables - **Benchmark command**: Runs with VRAM segment configuration \ No newline at end of file diff --git a/deploy/cloud/pre-deployment/nixl/build_and_deploy.sh b/deploy/cloud/pre-deployment/nixl/build_and_deploy.sh index 88f966a61bad..fa125f8cd56f 100755 --- a/deploy/cloud/pre-deployment/nixl/build_and_deploy.sh +++ b/deploy/cloud/pre-deployment/nixl/build_and_deploy.sh @@ -160,7 +160,7 @@ validate_architecture() { # Function to prompt for registry prompt_for_registry() { echo - printf "Enter your Docker registry (e.g., my-registry, docker.io/username): " + printf "Enter your Docker registry (e.g., nvcr.io/nvidia/ai-dynamo, docker.io/username): " read REGISTRY if [ -z "$REGISTRY" ]; then echo "Error: Registry cannot be empty" @@ -205,7 +205,7 @@ update_deployment() { cp "${SCRIPT_DIR}/nixlbench-deployment.yaml" "$deployment_file" # Update the image field using sed - sed -i "s|my-registry/nixlbench:version-arch|${registry}/nixlbench:${NIXL_VERSION}-${arch}|g" "$deployment_file" + sed -i "s|nvcr.io/nvidia/ai-dynamo/nixlbench:version-arch|${registry}/nixlbench:${NIXL_VERSION}-${arch}|g" "$deployment_file" echo "Deployment file updated with image: ${registry}/nixlbench:${NIXL_VERSION}-${arch}" } diff --git a/deploy/cloud/pre-deployment/nixl/nixlbench-deployment.yaml b/deploy/cloud/pre-deployment/nixl/nixlbench-deployment.yaml index 15cd39431555..811735fe0bec 100644 --- a/deploy/cloud/pre-deployment/nixl/nixlbench-deployment.yaml +++ b/deploy/cloud/pre-deployment/nixl/nixlbench-deployment.yaml @@ -16,7 +16,7 @@ spec: spec: containers: - name: nixl-benchmark - image: "my-registry/nixlbench:version-arch" + image: "nvcr.io/nvidia/ai-dynamo/nixlbench:version-arch" command: ["sh", "-c"] env: - name: ETCD_ENDPOINTS diff --git a/docs/README.md b/docs/README.md index 9d1f30bbb225..a7b987293241 100644 --- a/docs/README.md +++ b/docs/README.md @@ -68,6 +68,12 @@ This approach is ideal for CI/CD pipelines or when you want complete isolation f - `docs/_extensions/` - Custom Sphinx extensions - `docs/build/` - Generated documentation output (not tracked in git) +## Redirect Creation + +When moving or renaming files a redirect must be created. + +Redirect entries should be added to the `redirects` dictionary in `conf.py`. For detailed information on redirect syntax, see the [sphinx-reredirects usage documentation](https://documatt.com/sphinx-reredirects/usage/#introduction). + ## Dependency Management Documentation dependencies are defined in `pyproject.toml` under the `[dependency-groups]` section: diff --git a/docs/_includes/install.rst b/docs/_includes/install.rst index 3403c6f827b9..ab774e802ee8 100644 --- a/docs/_includes/install.rst +++ b/docs/_includes/install.rst @@ -10,7 +10,7 @@ Install a pre-built wheel from PyPI. source venv/bin/activate # Install Dynamo from PyPI (choose one backend extra) - uv pip install "ai-dynamo[sglang]==my-tag" # or [vllm], [trtllm] + uv pip install "ai-dynamo[sglang]==0.6.1" # or [vllm], [trtllm] Pip from source @@ -41,4 +41,4 @@ Pull and run prebuilt images from NVIDIA NGC (`nvcr.io`). docker run --rm -it \ --gpus all \ --network host \ - nvcr.io/nvidia/ai-dynamo/sglang-runtime:my-tag # or vllm, tensorrtllm + nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 # or vllm, tensorrtllm diff --git a/docs/backends/trtllm/gpt-oss.md b/docs/backends/trtllm/gpt-oss.md index 279f2fd1c280..057802a9faf8 100644 --- a/docs/backends/trtllm/gpt-oss.md +++ b/docs/backends/trtllm/gpt-oss.md @@ -49,7 +49,7 @@ huggingface-cli download openai/gpt-oss-120b --exclude "original/*" --exclude "m Set the container image: ```bash -export DYNAMO_CONTAINER_IMAGE=nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:my-tag +export DYNAMO_CONTAINER_IMAGE=nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:0.6.1 ``` Launch the Dynamo TensorRT-LLM container with the necessary configurations: diff --git a/docs/benchmarks/benchmarking.md b/docs/benchmarks/benchmarking.md index a981ecb4b6a4..2ee82dd73b72 100644 --- a/docs/benchmarks/benchmarking.md +++ b/docs/benchmarks/benchmarking.md @@ -410,7 +410,7 @@ The benchmark job is configured directly in the YAML file. - **Model**: `Qwen/Qwen3-0.6B` - **Benchmark Name**: `qwen3-0p6b-vllm-agg` - **Service**: `vllm-agg-frontend:8000` -- **Docker Image**: `nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag` +- **Docker Image**: `nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1` ### Customizing the Job diff --git a/docs/conf.py b/docs/conf.py index b4f50c9ed9b2..207a4ff84700 100644 --- a/docs/conf.py +++ b/docs/conf.py @@ -32,6 +32,23 @@ "sphinxcontrib.mermaid", ] +# Redirects configuration +redirects = { + "guides/tool-calling": "../agents/tool-calling.html", # key format corrected + "architecture/architecture": "../design_docs/architecture.html", + "architecture/disagg_serving": "../design_docs/disagg_serving.html", + "architecture/distributed_runtime": "../design_docs/distributed_runtime.html", + "architecture/dynamo_flow": "../design_docs/dynamo_flow.html", + "architecture/request_cancellation": "../fault_tolerance/request_cancellation.html", + "architecture/request_migration": "../fault_tolerance/request_migration.html", + "kubernetes/create_deployment": "../kubernetes/deployment/create_deployment.html", + "kubernetes/minikube": "../kubernetes/deployment/minikube.html", + "kubernetes/multinode-deployment": "../kubernetes/deployment/multinode-deployment.html", + "kubernetes/logging": "../kubernetes/observability/logging.html", + "kubernetes/metrics": "../kubernetes/observability/metrics.html", + "architecture/kv_cache_routing": "../router/kv_cache_routing.html", +} + # Custom extensions sys.path.insert(0, os.path.abspath("_extensions")) extensions.append("github_alerts") diff --git a/docs/kubernetes/deployment/create_deployment.md b/docs/kubernetes/deployment/create_deployment.md index a0d2877d0a64..54e94a30c91b 100644 --- a/docs/kubernetes/deployment/create_deployment.md +++ b/docs/kubernetes/deployment/create_deployment.md @@ -212,7 +212,7 @@ When disabled, you can manually specify secrets as you would for a normal pod sp nvidia.com/disable-image-pull-secret-discovery: "true" extraPodSpec: imagePullSecrets: - - name: my-registry-secret + - name: nvcr.io/nvidia/ai-dynamo-secret - name: another-secret mainContainer: image: your-image diff --git a/examples/basics/kubernetes/Distributed_Inference/README.md b/examples/basics/kubernetes/Distributed_Inference/README.md index d278ca41f35c..e4f5ebee7ec0 100644 --- a/examples/basics/kubernetes/Distributed_Inference/README.md +++ b/examples/basics/kubernetes/Distributed_Inference/README.md @@ -69,7 +69,7 @@ aiconfigurator cli --model LLAMA3.1_70B --total_gpus 16 --system h200_sxm ``` and from the output, you can see the Pareto curve with suggest P/D settings ![text](images/pareto.png) -3. Start the serving with 1 prefill worker with tensor parallelism 4 and 1 decoding worker with tensor parallelism 8 as AI Configurator suggested. Update the `my-tag` in `disagg_router.yaml` with the latest Dynamo version and your local cache folder path and run following command. +3. Start the serving with 1 prefill worker with tensor parallelism 4 and 1 decoding worker with tensor parallelism 8 as AI Configurator suggested. Update the `0.6.1` in `disagg_router.yaml` with the latest Dynamo version and your local cache folder path and run following command. ![text](images/settings.png) ```sh kubectl apply -f disagg_router.yaml --namespace ${NAMESPACE} diff --git a/examples/basics/kubernetes/Distributed_Inference/agg_router.yaml b/examples/basics/kubernetes/Distributed_Inference/agg_router.yaml index e05d2eab7645..744b841987e0 100644 --- a/examples/basics/kubernetes/Distributed_Inference/agg_router.yaml +++ b/examples/basics/kubernetes/Distributed_Inference/agg_router.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 envs: - name: DYN_ROUTER_MODE value: kv @@ -35,7 +35,7 @@ spec: path: /YOUR/LOCAL/CACHE/FOLDER type: DirectoryOrCreate mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 volumeMounts: - name: local-model-cache mountPath: /root/.cache diff --git a/examples/basics/kubernetes/Distributed_Inference/disagg_router.yaml b/examples/basics/kubernetes/Distributed_Inference/disagg_router.yaml index cc2e16aaf409..c3b58915c342 100644 --- a/examples/basics/kubernetes/Distributed_Inference/disagg_router.yaml +++ b/examples/basics/kubernetes/Distributed_Inference/disagg_router.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 envs: - name: DYN_ROUTER_MODE value: kv @@ -35,7 +35,7 @@ spec: path: /YOUR/LOCAL/CACHE/FOLDER type: DirectoryOrCreate mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm volumeMounts: - name: local-model-cache @@ -63,7 +63,7 @@ spec: path: /YOUR/LOCAL/CACHE/FOLDER type: DirectoryOrCreate mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm volumeMounts: - name: local-model-cache diff --git a/examples/custom_backend/hello_world/deploy/hello_world.yaml b/examples/custom_backend/hello_world/deploy/hello_world.yaml index 0fb8f4238ee5..8ded11a73b07 100644 --- a/examples/custom_backend/hello_world/deploy/hello_world.yaml +++ b/examples/custom_backend/hello_world/deploy/hello_world.yaml @@ -41,7 +41,7 @@ spec: memory: "2Gi" extraPodSpec: mainContainer: - image: my-registry/dynamo:my-tag + image: nvcr.io/nvidia/ai-dynamo/dynamo:0.6.1 workingDir: /workspace/examples/custom_backend/hello_world/ command: - /bin/sh @@ -80,7 +80,7 @@ spec: memory: "4Gi" extraPodSpec: mainContainer: - image: my-registry/dynamo:my-tag + image: nvcr.io/nvidia/ai-dynamo/dynamo:0.6.1 workingDir: /workspace/examples/custom_backend/hello_world/ command: - /bin/sh diff --git a/examples/deployments/ECS/task_definition_frontend.json b/examples/deployments/ECS/task_definition_frontend.json index fda0a2887669..cd83f62cadbd 100644 --- a/examples/deployments/ECS/task_definition_frontend.json +++ b/examples/deployments/ECS/task_definition_frontend.json @@ -3,7 +3,7 @@ "containerDefinitions": [ { "name": "dynamo-vllm-frontend", - "image": "nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag", + "image": "nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1", "repositoryCredentials": { "credentialsParameter": "arn:aws:secretsmanager:AWS_REGION:AWS_ID:secret:ngc_nvcr_access" }, diff --git a/examples/deployments/ECS/task_definition_prefillworker.json b/examples/deployments/ECS/task_definition_prefillworker.json index fdf928bf3d0b..01b113ae63fb 100644 --- a/examples/deployments/ECS/task_definition_prefillworker.json +++ b/examples/deployments/ECS/task_definition_prefillworker.json @@ -3,7 +3,7 @@ "containerDefinitions": [ { "name": "dynamo-prefill", - "image": "nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag", + "image": "nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1", "repositoryCredentials": { "credentialsParameter": "arn:aws:secretsmanager:AWS_REGION:AWS_ID:secret:ngc_access" }, diff --git a/examples/deployments/GKE/sglang/disagg.yaml b/examples/deployments/GKE/sglang/disagg.yaml index d64752e5698d..19ccfe74e3d7 100644 --- a/examples/deployments/GKE/sglang/disagg.yaml +++ b/examples/deployments/GKE/sglang/disagg.yaml @@ -12,7 +12,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 decode: envFromSecret: hf-token-secret dynamoNamespace: sglang-disagg @@ -24,7 +24,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 workingDir: /workspace/components/backends/sglang command: - /bin/sh @@ -47,7 +47,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/sglang-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-runtime:0.6.1 workingDir: /workspace/components/backends/sglang command: - /bin/sh diff --git a/examples/deployments/GKE/vllm/disagg.yaml b/examples/deployments/GKE/vllm/disagg.yaml index 098fe06d14b2..b525001605d2 100644 --- a/examples/deployments/GKE/vllm/disagg.yaml +++ b/examples/deployments/GKE/vllm/disagg.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 VllmDecodeWorker: dynamoNamespace: vllm-disagg envFromSecret: hf-token-secret @@ -27,7 +27,7 @@ spec: mainContainer: startupProbe: initialDelaySeconds: 180 - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - /bin/sh @@ -49,7 +49,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - /bin/sh diff --git a/examples/multimodal/deploy/agg_llava.yaml b/examples/multimodal/deploy/agg_llava.yaml index 8c125666b19b..820923f3388f 100644 --- a/examples/multimodal/deploy/agg_llava.yaml +++ b/examples/multimodal/deploy/agg_llava.yaml @@ -14,7 +14,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 EncodeWorker: envFromSecret: hf-token-secret dynamoNamespace: agg-llava @@ -25,7 +25,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/examples/multimodal command: - /bin/sh @@ -42,7 +42,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/examples/multimodal command: - /bin/sh @@ -59,7 +59,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/examples/multimodal command: - /bin/sh diff --git a/examples/multimodal/deploy/agg_qwen.yaml b/examples/multimodal/deploy/agg_qwen.yaml index 174979109a1b..babcd75dfb52 100644 --- a/examples/multimodal/deploy/agg_qwen.yaml +++ b/examples/multimodal/deploy/agg_qwen.yaml @@ -14,7 +14,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 EncodeWorker: envFromSecret: hf-token-secret dynamoNamespace: agg-qwen @@ -25,7 +25,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/examples/multimodal command: - /bin/sh @@ -42,7 +42,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/examples/multimodal command: - /bin/sh @@ -59,7 +59,7 @@ spec: gpu: "1" extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/examples/multimodal command: - /bin/sh diff --git a/recipes/deepseek-r1/sglang/disagg-16gpu/deploy.yaml b/recipes/deepseek-r1/sglang/disagg-16gpu/deploy.yaml index 888862b92f0a..040797cc500a 100644 --- a/recipes/deepseek-r1/sglang/disagg-16gpu/deploy.yaml +++ b/recipes/deepseek-r1/sglang/disagg-16gpu/deploy.yaml @@ -23,7 +23,7 @@ spec: periodSeconds: 10 timeoutSeconds: 1800 failureThreshold: 60 - image: my-registry/sglang-wideep-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-wideep-runtime:0.6.1 decode: dynamoNamespace: sgl-dsr1-16gpu componentType: worker @@ -47,7 +47,7 @@ spec: periodSeconds: 10 timeoutSeconds: 10 failureThreshold: 600 - image: my-registry/sglang-wideep-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-wideep-runtime:0.6.1 workingDir: /workspace/components/backends/sglang command: - python3 @@ -100,7 +100,7 @@ spec: periodSeconds: 10 timeoutSeconds: 10 failureThreshold: 600 - image: my-registry/sglang-wideep-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-wideep-runtime:0.6.1 workingDir: /workspace/components/backends/sglang command: - python3 diff --git a/recipes/deepseek-r1/sglang/disagg-8gpu/deploy.yaml b/recipes/deepseek-r1/sglang/disagg-8gpu/deploy.yaml index b1f96e592e8e..1fc945764ab2 100644 --- a/recipes/deepseek-r1/sglang/disagg-8gpu/deploy.yaml +++ b/recipes/deepseek-r1/sglang/disagg-8gpu/deploy.yaml @@ -23,7 +23,7 @@ spec: periodSeconds: 10 timeoutSeconds: 1800 failureThreshold: 60 - image: my-registry/sglang-wideep-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-wideep-runtime:0.6.1 decode: dynamoNamespace: sgl-dsr1-8gpu componentType: worker @@ -45,7 +45,7 @@ spec: periodSeconds: 10 timeoutSeconds: 10 failureThreshold: 600 - image: my-registry/sglang-wideep-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-wideep-runtime:0.6.1 workingDir: /workspace/components/backends/sglang command: - python3 @@ -94,7 +94,7 @@ spec: periodSeconds: 10 timeoutSeconds: 10 failureThreshold: 600 - image: my-registry/sglang-wideep-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/sglang-wideep-runtime:0.6.1 workingDir: /workspace/components/backends/sglang command: - python3 diff --git a/recipes/gpt-oss-120b/trtllm/agg/deploy.yaml b/recipes/gpt-oss-120b/trtllm/agg/deploy.yaml index a3f0b5c2e636..83dd90eb2bec 100644 --- a/recipes/gpt-oss-120b/trtllm/agg/deploy.yaml +++ b/recipes/gpt-oss-120b/trtllm/agg/deploy.yaml @@ -46,7 +46,7 @@ spec: command: - /bin/sh - -c - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 replicas: 1 TrtllmWorker: componentType: main @@ -81,7 +81,7 @@ spec: command: - /bin/sh - -c - image: my-registry/trtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/trtllm-runtime:0.6.1 env: - name: TRTLLM_ENABLE_PDL value: "1" diff --git a/recipes/llama-3-70b/vllm/agg/deploy.yaml b/recipes/llama-3-70b/vllm/agg/deploy.yaml index 1a0101e26f8d..7eace2535367 100644 --- a/recipes/llama-3-70b/vllm/agg/deploy.yaml +++ b/recipes/llama-3-70b/vllm/agg/deploy.yaml @@ -18,7 +18,7 @@ spec: mountPoint: /root/.cache/huggingface extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm replicas: 1 VllmPrefillWorker: @@ -42,7 +42,7 @@ spec: command: - /bin/sh - -c - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm replicas: 1 resources: diff --git a/recipes/llama-3-70b/vllm/disagg-multi-node/deploy.yaml b/recipes/llama-3-70b/vllm/disagg-multi-node/deploy.yaml index 229e7196c326..1f606d686cec 100644 --- a/recipes/llama-3-70b/vllm/disagg-multi-node/deploy.yaml +++ b/recipes/llama-3-70b/vllm/disagg-multi-node/deploy.yaml @@ -18,7 +18,7 @@ spec: mountPoint: /root/.cache/huggingface extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm replicas: 1 VllmPrefillWorker: @@ -42,7 +42,7 @@ spec: command: - /bin/sh - -c - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm replicas: 1 resources: @@ -71,7 +71,7 @@ spec: command: - /bin/sh - -c - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm replicas: 1 resources: diff --git a/recipes/llama-3-70b/vllm/disagg-single-node/deploy.yaml b/recipes/llama-3-70b/vllm/disagg-single-node/deploy.yaml index fd2c16fb6212..972ea77f06c4 100644 --- a/recipes/llama-3-70b/vllm/disagg-single-node/deploy.yaml +++ b/recipes/llama-3-70b/vllm/disagg-single-node/deploy.yaml @@ -18,7 +18,7 @@ spec: mountPoint: /root/.cache/huggingface extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm replicas: 1 VllmPrefillWorker: @@ -52,7 +52,7 @@ spec: command: - /bin/sh - -c - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm replicas: 2 resources: @@ -91,7 +91,7 @@ spec: command: - /bin/sh - -c - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm replicas: 1 resources: diff --git a/recipes/qwen3-32b-fp8/trtllm/agg/deploy.yaml b/recipes/qwen3-32b-fp8/trtllm/agg/deploy.yaml index b11aa10c6b78..3f0814c81d4f 100644 --- a/recipes/qwen3-32b-fp8/trtllm/agg/deploy.yaml +++ b/recipes/qwen3-32b-fp8/trtllm/agg/deploy.yaml @@ -62,7 +62,7 @@ spec: command: - /bin/sh - -c - image: nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:0.6.1 replicas: 1 TrtllmWorker: componentType: main @@ -96,7 +96,7 @@ spec: command: - /bin/sh - -c - image: nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:0.6.1 env: - name: TRTLLM_ENABLE_PDL value: "1" diff --git a/recipes/qwen3-32b-fp8/trtllm/disagg/deploy.yaml b/recipes/qwen3-32b-fp8/trtllm/disagg/deploy.yaml index 6fb79287e190..4834c3b70861 100644 --- a/recipes/qwen3-32b-fp8/trtllm/disagg/deploy.yaml +++ b/recipes/qwen3-32b-fp8/trtllm/disagg/deploy.yaml @@ -219,7 +219,7 @@ spec: command: - /bin/sh - -c - image: nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:0.6.1 replicas: 1 TrtllmPrefillWorker: componentType: worker @@ -256,7 +256,7 @@ spec: command: - /bin/sh - -c - image: nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:0.6.1 env: - name: TRTLLM_ENABLE_PDL value: "1" @@ -316,7 +316,7 @@ spec: command: - /bin/sh - -c - image: nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:0.6.1 env: - name: TRTLLM_ENABLE_PDL value: "1" diff --git a/tests/fault_tolerance/deploy/templates/vllm/moe_agg.yaml b/tests/fault_tolerance/deploy/templates/vllm/moe_agg.yaml index 3e9c58926c22..dd9cf8c5b4e4 100644 --- a/tests/fault_tolerance/deploy/templates/vllm/moe_agg.yaml +++ b/tests/fault_tolerance/deploy/templates/vllm/moe_agg.yaml @@ -13,7 +13,7 @@ spec: replicas: 1 extraPodSpec: mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 VllmDecodeWorker: envFromSecret: hf-token-secret dynamoNamespace: vllm-moe-agg @@ -48,7 +48,7 @@ spec: imagePullSecrets: - name: nvcr-imagepullsecret mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 diff --git a/tests/fault_tolerance/deploy/templates/vllm/moe_disagg.yaml b/tests/fault_tolerance/deploy/templates/vllm/moe_disagg.yaml index b60018829c7c..47d689f47edd 100644 --- a/tests/fault_tolerance/deploy/templates/vllm/moe_disagg.yaml +++ b/tests/fault_tolerance/deploy/templates/vllm/moe_disagg.yaml @@ -15,7 +15,7 @@ spec: imagePullSecrets: - name: nvcr-imagepullsecret mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 VllmDecodeWorker: dynamoNamespace: vllm-moe-disagg envFromSecret: hf-token-secret @@ -51,7 +51,7 @@ spec: imagePullSecrets: - name: nvcr-imagepullsecret mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 @@ -116,7 +116,7 @@ spec: imagePullSecrets: - name: nvcr-imagepullsecret mainContainer: - image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 diff --git a/tests/planner/perf_test_configs/agg_8b.yaml b/tests/planner/perf_test_configs/agg_8b.yaml index c176e62cec71..1676fd18a25d 100644 --- a/tests/planner/perf_test_configs/agg_8b.yaml +++ b/tests/planner/perf_test_configs/agg_8b.yaml @@ -38,7 +38,7 @@ spec: memory: "100Gi" extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - /bin/sh @@ -88,7 +88,7 @@ spec: port: 9090 periodSeconds: 10 failureThreshold: 60 - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - /bin/sh diff --git a/tests/planner/perf_test_configs/disagg_8b_2p2d.yaml b/tests/planner/perf_test_configs/disagg_8b_2p2d.yaml index 4b38a9a9b048..59e0bd85df39 100644 --- a/tests/planner/perf_test_configs/disagg_8b_2p2d.yaml +++ b/tests/planner/perf_test_configs/disagg_8b_2p2d.yaml @@ -38,7 +38,7 @@ spec: memory: "100Gi" extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - /bin/sh @@ -88,7 +88,7 @@ spec: port: 9090 periodSeconds: 10 failureThreshold: 60 - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - /bin/sh @@ -138,7 +138,7 @@ spec: port: 9090 periodSeconds: 10 failureThreshold: 60 - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - /bin/sh diff --git a/tests/planner/perf_test_configs/disagg_8b_3p1d.yaml b/tests/planner/perf_test_configs/disagg_8b_3p1d.yaml index 41618cf472a1..cc9ba08dcca7 100644 --- a/tests/planner/perf_test_configs/disagg_8b_3p1d.yaml +++ b/tests/planner/perf_test_configs/disagg_8b_3p1d.yaml @@ -38,7 +38,7 @@ spec: memory: "100Gi" extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - /bin/sh @@ -88,7 +88,7 @@ spec: port: 9090 periodSeconds: 10 failureThreshold: 60 - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - /bin/sh @@ -138,7 +138,7 @@ spec: port: 9090 periodSeconds: 10 failureThreshold: 60 - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - /bin/sh diff --git a/tests/planner/perf_test_configs/disagg_8b_planner.yaml b/tests/planner/perf_test_configs/disagg_8b_planner.yaml index d3e1761fa2bc..35212cb9a330 100644 --- a/tests/planner/perf_test_configs/disagg_8b_planner.yaml +++ b/tests/planner/perf_test_configs/disagg_8b_planner.yaml @@ -41,7 +41,7 @@ spec: memory: "100Gi" extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - /bin/sh @@ -74,7 +74,7 @@ spec: failureThreshold: 10 extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/src/dynamo/planner ports: - name: metrics @@ -138,7 +138,7 @@ spec: port: 9090 periodSeconds: 10 failureThreshold: 60 - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 @@ -195,7 +195,7 @@ spec: port: 9090 periodSeconds: 10 failureThreshold: 60 - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - python3 diff --git a/tests/planner/perf_test_configs/disagg_8b_tp2.yaml b/tests/planner/perf_test_configs/disagg_8b_tp2.yaml index 3c83f78bcda6..87f20f305f6f 100644 --- a/tests/planner/perf_test_configs/disagg_8b_tp2.yaml +++ b/tests/planner/perf_test_configs/disagg_8b_tp2.yaml @@ -38,7 +38,7 @@ spec: memory: "100Gi" extraPodSpec: mainContainer: - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - /bin/sh @@ -88,7 +88,7 @@ spec: port: 9090 periodSeconds: 10 failureThreshold: 60 - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - /bin/sh @@ -138,7 +138,7 @@ spec: port: 9090 periodSeconds: 10 failureThreshold: 60 - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 workingDir: /workspace/components/backends/vllm command: - /bin/sh diff --git a/tests/planner/perf_test_configs/image_cache_daemonset.yaml b/tests/planner/perf_test_configs/image_cache_daemonset.yaml index 026337e5bab1..434220c0b6be 100644 --- a/tests/planner/perf_test_configs/image_cache_daemonset.yaml +++ b/tests/planner/perf_test_configs/image_cache_daemonset.yaml @@ -20,7 +20,7 @@ spec: - name: nvcr-imagepullsecret containers: - name: image-cache - image: my-registry/vllm-runtime:my-tag + image: nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.6.1 command: - /bin/sh - -c