biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[release] Move agent queue to Release cluster queues (#37783)

Browse Source 
Signed-off-by: khluu <khluu000@gmail.com>
(cherry picked from commit 7281199a8c)
This commit is contained in:
Kevin H. Luu
2026-03-23 20:36:47 -07:00
committed by khluu
parent 90b29e5302
commit b7e4b88987
1 changed files with 24 additions and 24 deletions
Download Patch File Download Diff File Expand all files Collapse all files

48
.buildkite/release-pipeline.yaml
View File

@@ -12,7 +12,7 @@ steps:
depends_on: ~ depends_on: ~
id: build-wheel-arm64-cuda-12-9 id: build-wheel-arm64-cuda-12-9
agents: agents:
queue: arm64_cpu_queue_postmerge queue: arm64_cpu_queue_release
commands: commands:
# #NOTE: torch_cuda_arch_list is derived from upstream PyTorch build files here: # #NOTE: torch_cuda_arch_list is derived from upstream PyTorch build files here:
# https://github.com/pytorch/pytorch/blob/main/.ci/aarch64_linux/aarch64_ci_build.sh#L7 # https://github.com/pytorch/pytorch/blob/main/.ci/aarch64_linux/aarch64_ci_build.sh#L7
@@ -27,7 +27,7 @@ steps:
depends_on: ~ depends_on: ~
id: build-wheel-arm64-cuda-13-0 id: build-wheel-arm64-cuda-13-0
agents: agents:
queue: arm64_cpu_queue_postmerge queue: arm64_cpu_queue_release
commands: commands:
# #NOTE: torch_cuda_arch_list is derived from upstream PyTorch build files here: # #NOTE: torch_cuda_arch_list is derived from upstream PyTorch build files here:
# https://github.com/pytorch/pytorch/blob/main/.ci/aarch64_linux/aarch64_ci_build.sh#L7 # https://github.com/pytorch/pytorch/blob/main/.ci/aarch64_linux/aarch64_ci_build.sh#L7
@@ -42,7 +42,7 @@ steps:
depends_on: ~ depends_on: ~
id: build-wheel-arm64-cpu id: build-wheel-arm64-cpu
agents: agents:
queue: arm64_cpu_queue_postmerge queue: arm64_cpu_queue_release
commands: commands:
- "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg GIT_REPO_CHECK=1 --build-arg VLLM_BUILD_ACL=ON --tag vllm-ci:build-image --target vllm-build --progress plain -f docker/Dockerfile.cpu ." - "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg GIT_REPO_CHECK=1 --build-arg VLLM_BUILD_ACL=ON --tag vllm-ci:build-image --target vllm-build --progress plain -f docker/Dockerfile.cpu ."
- "mkdir artifacts" - "mkdir artifacts"
@@ -55,7 +55,7 @@ steps:
depends_on: ~ depends_on: ~
id: build-wheel-x86-cuda-12-9 id: build-wheel-x86-cuda-12-9
agents: agents:
queue: cpu_queue_postmerge queue: cpu_queue_release
commands: commands:
- "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg USE_SCCACHE=1 --build-arg GIT_REPO_CHECK=1 --build-arg CUDA_VERSION=12.9.1 --tag vllm-ci:build-image --target build --progress plain -f docker/Dockerfile ." - "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg USE_SCCACHE=1 --build-arg GIT_REPO_CHECK=1 --build-arg CUDA_VERSION=12.9.1 --tag vllm-ci:build-image --target build --progress plain -f docker/Dockerfile ."
- "mkdir artifacts" - "mkdir artifacts"
@@ -68,7 +68,7 @@ steps:
depends_on: ~ depends_on: ~
id: build-wheel-x86-cuda-13-0 id: build-wheel-x86-cuda-13-0
agents: agents:
queue: cpu_queue_postmerge queue: cpu_queue_release
commands: commands:
- "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg USE_SCCACHE=1 --build-arg GIT_REPO_CHECK=1 --build-arg CUDA_VERSION=13.0.1 --build-arg BUILD_BASE_IMAGE=nvidia/cuda:13.0.1-devel-ubuntu22.04 --tag vllm-ci:build-image --target build --progress plain -f docker/Dockerfile ." - "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg USE_SCCACHE=1 --build-arg GIT_REPO_CHECK=1 --build-arg CUDA_VERSION=13.0.1 --build-arg BUILD_BASE_IMAGE=nvidia/cuda:13.0.1-devel-ubuntu22.04 --tag vllm-ci:build-image --target build --progress plain -f docker/Dockerfile ."
- "mkdir artifacts" - "mkdir artifacts"
@@ -81,7 +81,7 @@ steps:
depends_on: ~ depends_on: ~
id: build-wheel-x86-cpu id: build-wheel-x86-cpu
agents: agents:
queue: cpu_queue_postmerge queue: cpu_queue_release
commands: commands:
- "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg GIT_REPO_CHECK=1 --build-arg VLLM_CPU_X86=true --tag vllm-ci:build-image --target vllm-build --progress plain -f docker/Dockerfile.cpu ." - "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg GIT_REPO_CHECK=1 --build-arg VLLM_CPU_X86=true --tag vllm-ci:build-image --target vllm-build --progress plain -f docker/Dockerfile.cpu ."
- "mkdir artifacts" - "mkdir artifacts"
@@ -97,7 +97,7 @@ steps:
depends_on: ~ depends_on: ~
id: build-release-image-x86 id: build-release-image-x86
agents: agents:
queue: cpu_queue_postmerge queue: cpu_queue_release
commands: commands:
- "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7" - "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7"
- "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg USE_SCCACHE=1 --build-arg GIT_REPO_CHECK=1 --build-arg CUDA_VERSION=12.9.1 --build-arg FLASHINFER_AOT_COMPILE=true --build-arg INSTALL_KV_CONNECTORS=true --tag public.ecr.aws/q9t5s3a7/vllm-release-repo:$BUILDKITE_COMMIT-$(uname -m) --target vllm-openai --progress plain -f docker/Dockerfile ." - "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg USE_SCCACHE=1 --build-arg GIT_REPO_CHECK=1 --build-arg CUDA_VERSION=12.9.1 --build-arg FLASHINFER_AOT_COMPILE=true --build-arg INSTALL_KV_CONNECTORS=true --tag public.ecr.aws/q9t5s3a7/vllm-release-repo:$BUILDKITE_COMMIT-$(uname -m) --target vllm-openai --progress plain -f docker/Dockerfile ."
@@ -110,7 +110,7 @@ steps:
depends_on: ~ depends_on: ~
id: build-release-image-arm64 id: build-release-image-arm64
agents: agents:
queue: arm64_cpu_queue_postmerge queue: arm64_cpu_queue_release
commands: commands:
- "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7" - "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7"
- "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg USE_SCCACHE=1 --build-arg GIT_REPO_CHECK=1 --build-arg CUDA_VERSION=12.9.1 --build-arg FLASHINFER_AOT_COMPILE=true --build-arg torch_cuda_arch_list='8.7 8.9 9.0 10.0+PTX 12.0' --build-arg INSTALL_KV_CONNECTORS=true --tag public.ecr.aws/q9t5s3a7/vllm-release-repo:$BUILDKITE_COMMIT-$(uname -m) --target vllm-openai --progress plain -f docker/Dockerfile ." - "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg USE_SCCACHE=1 --build-arg GIT_REPO_CHECK=1 --build-arg CUDA_VERSION=12.9.1 --build-arg FLASHINFER_AOT_COMPILE=true --build-arg torch_cuda_arch_list='8.7 8.9 9.0 10.0+PTX 12.0' --build-arg INSTALL_KV_CONNECTORS=true --tag public.ecr.aws/q9t5s3a7/vllm-release-repo:$BUILDKITE_COMMIT-$(uname -m) --target vllm-openai --progress plain -f docker/Dockerfile ."
@@ -120,7 +120,7 @@ steps:
depends_on: ~ depends_on: ~
id: build-release-image-x86-cuda-13-0 id: build-release-image-x86-cuda-13-0
agents: agents:
queue: cpu_queue_postmerge queue: cpu_queue_release
commands: commands:
- "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7" - "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7"
- "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg USE_SCCACHE=1 --build-arg GIT_REPO_CHECK=1 --build-arg CUDA_VERSION=13.0.1 --build-arg INSTALL_KV_CONNECTORS=true --build-arg BUILD_BASE_IMAGE=nvidia/cuda:13.0.1-devel-ubuntu22.04 --tag public.ecr.aws/q9t5s3a7/vllm-release-repo:$BUILDKITE_COMMIT-$(uname -m)-cu130 --target vllm-openai --progress plain -f docker/Dockerfile ." - "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg USE_SCCACHE=1 --build-arg GIT_REPO_CHECK=1 --build-arg CUDA_VERSION=13.0.1 --build-arg INSTALL_KV_CONNECTORS=true --build-arg BUILD_BASE_IMAGE=nvidia/cuda:13.0.1-devel-ubuntu22.04 --tag public.ecr.aws/q9t5s3a7/vllm-release-repo:$BUILDKITE_COMMIT-$(uname -m)-cu130 --target vllm-openai --progress plain -f docker/Dockerfile ."
@@ -133,7 +133,7 @@ steps:
depends_on: ~ depends_on: ~
id: build-release-image-arm64-cuda-13-0 id: build-release-image-arm64-cuda-13-0
agents: agents:
queue: arm64_cpu_queue_postmerge queue: arm64_cpu_queue_release
commands: commands:
- "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7" - "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7"
# compute capability 12.0 for RTX-50 series / RTX PRO 6000 Blackwell, 12.1 for DGX Spark # compute capability 12.0 for RTX-50 series / RTX PRO 6000 Blackwell, 12.1 for DGX Spark
@@ -149,7 +149,7 @@ steps:
- block-cpu-release-image-build - block-cpu-release-image-build
- input-release-version - input-release-version
agents: agents:
queue: cpu_queue_postmerge queue: cpu_queue_release
commands: commands:
- "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7" - "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7"
- "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg GIT_REPO_CHECK=1 --build-arg VLLM_CPU_X86=true --tag public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:$(buildkite-agent meta-data get release-version) --tag public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:latest --progress plain --target vllm-openai -f docker/Dockerfile.cpu ." - "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg GIT_REPO_CHECK=1 --build-arg VLLM_CPU_X86=true --tag public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:$(buildkite-agent meta-data get release-version) --tag public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:latest --progress plain --target vllm-openai -f docker/Dockerfile.cpu ."
@@ -167,7 +167,7 @@ steps:
- block-arm64-cpu-release-image-build - block-arm64-cpu-release-image-build
- input-release-version - input-release-version
agents: agents:
queue: arm64_cpu_queue_postmerge queue: arm64_cpu_queue_release
commands: commands:
- "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7" - "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7"
- "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg GIT_REPO_CHECK=1 --tag public.ecr.aws/q9t5s3a7/vllm-arm64-cpu-release-repo:$(buildkite-agent meta-data get release-version) --tag public.ecr.aws/q9t5s3a7/vllm-arm64-cpu-release-repo:latest --progress plain --target vllm-openai -f docker/Dockerfile.cpu ." - "DOCKER_BUILDKIT=1 docker build --build-arg max_jobs=16 --build-arg GIT_REPO_CHECK=1 --tag public.ecr.aws/q9t5s3a7/vllm-arm64-cpu-release-repo:$(buildkite-agent meta-data get release-version) --tag public.ecr.aws/q9t5s3a7/vllm-arm64-cpu-release-repo:latest --progress plain --target vllm-openai -f docker/Dockerfile.cpu ."
@@ -185,7 +185,7 @@ steps:
- build-release-image-arm64 - build-release-image-arm64
id: create-multi-arch-manifest id: create-multi-arch-manifest
agents: agents:
queue: small_cpu_queue_postmerge queue: small_cpu_queue_release
commands: commands:
- "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7" - "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7"
- "docker manifest create public.ecr.aws/q9t5s3a7/vllm-release-repo:$BUILDKITE_COMMIT public.ecr.aws/q9t5s3a7/vllm-release-repo:$BUILDKITE_COMMIT-x86_64 public.ecr.aws/q9t5s3a7/vllm-release-repo:$BUILDKITE_COMMIT-aarch64 --amend" - "docker manifest create public.ecr.aws/q9t5s3a7/vllm-release-repo:$BUILDKITE_COMMIT public.ecr.aws/q9t5s3a7/vllm-release-repo:$BUILDKITE_COMMIT-x86_64 public.ecr.aws/q9t5s3a7/vllm-release-repo:$BUILDKITE_COMMIT-aarch64 --amend"
@@ -196,7 +196,7 @@ steps:
- create-multi-arch-manifest - create-multi-arch-manifest
id: annotate-release-workflow id: annotate-release-workflow
agents: agents:
queue: small_cpu_queue_postmerge queue: small_cpu_queue_release
commands: commands:
- "bash .buildkite/scripts/annotate-release.sh" - "bash .buildkite/scripts/annotate-release.sh"
@@ -206,7 +206,7 @@ steps:
- build-release-image-arm64-cuda-13-0 - build-release-image-arm64-cuda-13-0
id: create-multi-arch-manifest-cuda-13-0 id: create-multi-arch-manifest-cuda-13-0
agents: agents:
queue: small_cpu_queue_postmerge queue: small_cpu_queue_release
commands: commands:
- "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7" - "aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws/q9t5s3a7"
- "docker manifest create public.ecr.aws/q9t5s3a7/vllm-release-repo:$BUILDKITE_COMMIT-cu130 public.ecr.aws/q9t5s3a7/vllm-release-repo:$BUILDKITE_COMMIT-x86_64-cu130 public.ecr.aws/q9t5s3a7/vllm-release-repo:$BUILDKITE_COMMIT-aarch64-cu130 --amend" - "docker manifest create public.ecr.aws/q9t5s3a7/vllm-release-repo:$BUILDKITE_COMMIT-cu130 public.ecr.aws/q9t5s3a7/vllm-release-repo:$BUILDKITE_COMMIT-x86_64-cu130 public.ecr.aws/q9t5s3a7/vllm-release-repo:$BUILDKITE_COMMIT-aarch64-cu130 --amend"
@@ -217,7 +217,7 @@ steps:
- create-multi-arch-manifest - create-multi-arch-manifest
if: build.env("NIGHTLY") == "1" if: build.env("NIGHTLY") == "1"
agents: agents:
queue: small_cpu_queue_postmerge queue: small_cpu_queue_release
commands: commands:
- "bash .buildkite/scripts/push-nightly-builds.sh" - "bash .buildkite/scripts/push-nightly-builds.sh"
# Clean up old nightly builds (keep only last 14) # Clean up old nightly builds (keep only last 14)
@@ -235,7 +235,7 @@ steps:
- create-multi-arch-manifest-cuda-13-0 - create-multi-arch-manifest-cuda-13-0
if: build.env("NIGHTLY") == "1" if: build.env("NIGHTLY") == "1"
agents: agents:
queue: small_cpu_queue_postmerge queue: small_cpu_queue_release
commands: commands:
- "bash .buildkite/scripts/push-nightly-builds.sh cu130" - "bash .buildkite/scripts/push-nightly-builds.sh cu130"
# Clean up old nightly builds (keep only last 14) # Clean up old nightly builds (keep only last 14)
@@ -262,7 +262,7 @@ steps:
- block-upload-release-wheels - block-upload-release-wheels
id: upload-release-wheels id: upload-release-wheels
agents: agents:
queue: small_cpu_queue_postmerge queue: small_cpu_queue_release
commands: commands:
- "bash .buildkite/scripts/upload-release-wheels-pypi.sh" - "bash .buildkite/scripts/upload-release-wheels-pypi.sh"
@@ -323,7 +323,7 @@ steps:
- step: input-rocm-config - step: input-rocm-config
allow_failure: true # Allow failure so non-UI builds can proceed (input step is skipped) allow_failure: true # Allow failure so non-UI builds can proceed (input step is skipped)
agents: agents:
queue: cpu_queue_postmerge queue: cpu_queue_release
commands: commands:
# Set configuration and check cache # Set configuration and check cache
- | - |
@@ -465,7 +465,7 @@ steps:
- step: build-rocm-base-wheels - step: build-rocm-base-wheels
allow_failure: false allow_failure: false
agents: agents:
queue: cpu_queue_postmerge queue: cpu_queue_release
timeout_in_minutes: 180 timeout_in_minutes: 180
commands: commands:
# Download artifacts and prepare Docker image # Download artifacts and prepare Docker image
@@ -575,7 +575,7 @@ steps:
- step: build-rocm-vllm-wheel - step: build-rocm-vllm-wheel
allow_failure: false allow_failure: false
agents: agents:
queue: cpu_queue_postmerge queue: cpu_queue_release
timeout_in_minutes: 60 timeout_in_minutes: 60
commands: commands:
# Download all wheel artifacts and run upload # Download all wheel artifacts and run upload
@@ -624,7 +624,7 @@ steps:
- step: input-release-version - step: input-release-version
allow_failure: true allow_failure: true
agents: agents:
queue: cpu_queue_postmerge queue: cpu_queue_release
commands: commands:
- "bash .buildkite/scripts/annotate-rocm-release.sh" - "bash .buildkite/scripts/annotate-rocm-release.sh"
env: env:
@@ -641,7 +641,7 @@ steps:
depends_on: block-generate-root-index-rocm-wheels depends_on: block-generate-root-index-rocm-wheels
id: generate-root-index-rocm-wheels id: generate-root-index-rocm-wheels
agents: agents:
queue: cpu_queue_postmerge queue: cpu_queue_release
commands: commands:
- "bash tools/vllm-rocm/generate-rocm-wheels-root-index.sh" - "bash tools/vllm-rocm/generate-rocm-wheels-root-index.sh"
env: env:
@@ -655,7 +655,7 @@ steps:
- step: build-rocm-base-wheels - step: build-rocm-base-wheels
allow_failure: false allow_failure: false
agents: agents:
queue: cpu_queue_postmerge queue: cpu_queue_release
timeout_in_minutes: 60 timeout_in_minutes: 60
commands: commands:
- | - |