diff --git a/.github/workflows/release-docker-amd-miles-rocm700-nightly.yml b/.github/workflows/release-docker-amd-miles-rocm700-nightly.yml index ed75f5295..eee036118 100644 --- a/.github/workflows/release-docker-amd-miles-rocm700-nightly.yml +++ b/.github/workflows/release-docker-amd-miles-rocm700-nightly.yml @@ -59,53 +59,3 @@ jobs: -t rocm/sgl-dev:${dated} \ --no-cache docker push rocm/sgl-dev:${dated} - - - name: Save published image tag - run: | - mkdir -p image-tag - echo "${{ env.IMAGE_TAG }}" > "image-tag/${{ matrix.gpu_arch }}.txt" - - - name: Upload image tag artifact - uses: actions/upload-artifact@v4 - with: - name: image-tag-${{ matrix.gpu_arch }} - path: image-tag/${{ matrix.gpu_arch }}.txt - retention-days: 1 - - push_local_registry: - if: ${{ !cancelled() && github.repository == 'sgl-project/sglang' }} - runs-on: linux-mi300-1gpu-sglang - environment: "prod" - needs: publish - strategy: - fail-fast: false - matrix: - gpu_arch: ["gfx950", "gfx942"] - steps: - - name: Download image tag artifact - uses: actions/download-artifact@v4 - with: - name: image-tag-${{ matrix.gpu_arch }} - - - name: Read image tag - run: | - image_tag=$(tr -d "[:space:]" < "${{ matrix.gpu_arch }}.txt") - if [ -z "${image_tag}" ]; then - echo "::error::Image tag artifact is empty" - exit 1 - fi - echo "IMAGE_TAG=${image_tag}" >> $GITHUB_ENV - - - name: Login to Docker Hub (AMD) - uses: docker/login-action@v2 - with: - username: ${{ secrets.DOCKERHUB_AMD_USERNAME }} - password: ${{ secrets.DOCKERHUB_AMD_TOKEN }} - - - name: Mirror rocm/sgl-dev to local registry - run: | - src="rocm/sgl-dev:${{ env.IMAGE_TAG }}" - dst="10.245.143.50:5000/rocm/sgl-dev:${{ env.IMAGE_TAG }}" - docker pull "${src}" - docker tag "${src}" "${dst}" - docker push "${dst}" diff --git a/.github/workflows/release-docker-amd-miles-rocm720-nightly.yml b/.github/workflows/release-docker-amd-miles-rocm720-nightly.yml index 42de726e5..75044f332 100644 --- a/.github/workflows/release-docker-amd-miles-rocm720-nightly.yml +++ b/.github/workflows/release-docker-amd-miles-rocm720-nightly.yml @@ -56,53 +56,3 @@ jobs: -t rocm/sgl-dev:${dated} \ --no-cache docker push rocm/sgl-dev:${dated} - - - name: Save published image tag - run: | - mkdir -p image-tag - echo "${{ env.IMAGE_TAG }}" > "image-tag/${{ matrix.gpu_arch }}.txt" - - - name: Upload image tag artifact - uses: actions/upload-artifact@v4 - with: - name: image-tag-${{ matrix.gpu_arch }} - path: image-tag/${{ matrix.gpu_arch }}.txt - retention-days: 1 - - push_local_registry: - if: ${{ !cancelled() && github.repository == 'sgl-project/sglang' }} - runs-on: linux-mi300-1gpu-sglang - environment: "prod" - needs: publish - strategy: - fail-fast: false - matrix: - gpu_arch: ["gfx950"] - steps: - - name: Download image tag artifact - uses: actions/download-artifact@v4 - with: - name: image-tag-${{ matrix.gpu_arch }} - - - name: Read image tag - run: | - image_tag=$(tr -d "[:space:]" < "${{ matrix.gpu_arch }}.txt") - if [ -z "${image_tag}" ]; then - echo "::error::Image tag artifact is empty" - exit 1 - fi - echo "IMAGE_TAG=${image_tag}" >> $GITHUB_ENV - - - name: Login to Docker Hub (AMD) - uses: docker/login-action@v2 - with: - username: ${{ secrets.DOCKERHUB_AMD_USERNAME }} - password: ${{ secrets.DOCKERHUB_AMD_TOKEN }} - - - name: Mirror rocm/sgl-dev to local registry - run: | - src="rocm/sgl-dev:${{ env.IMAGE_TAG }}" - dst="10.245.143.50:5000/rocm/sgl-dev:${{ env.IMAGE_TAG }}" - docker pull "${src}" - docker tag "${src}" "${dst}" - docker push "${dst}" diff --git a/.github/workflows/release-docker-amd-nightly.yml b/.github/workflows/release-docker-amd-nightly.yml index db3d6ba2b..420c01d8d 100644 --- a/.github/workflows/release-docker-amd-nightly.yml +++ b/.github/workflows/release-docker-amd-nightly.yml @@ -158,7 +158,7 @@ jobs: - name: Mirror rocm/sgl-dev to local registry run: | src="rocm/sgl-dev:${{ env.IMAGE_TAG }}" - dst="10.245.143.50:5000/rocm/sgl-dev:${{ env.IMAGE_TAG }}" + dst="10.44.14.109:5000/rocm/sgl-dev:${{ env.IMAGE_TAG }}" docker pull "${src}" docker tag "${src}" "${dst}" docker push "${dst}" diff --git a/.github/workflows/release-docker-amd-rocm720-nightly.yml b/.github/workflows/release-docker-amd-rocm720-nightly.yml index 67eabdefd..fae14790a 100644 --- a/.github/workflows/release-docker-amd-rocm720-nightly.yml +++ b/.github/workflows/release-docker-amd-rocm720-nightly.yml @@ -167,7 +167,7 @@ jobs: - name: Mirror rocm/sgl-dev to local registry run: | src="rocm/sgl-dev:${{ env.IMAGE_TAG }}" - dst="10.245.143.50:5000/rocm/sgl-dev:${{ env.IMAGE_TAG }}" + dst="10.44.14.109:5000/rocm/sgl-dev:${{ env.IMAGE_TAG }}" docker pull "${src}" docker tag "${src}" "${dst}" docker push "${dst}" diff --git a/scripts/ci/amd/amd_ci_start_container.sh b/scripts/ci/amd/amd_ci_start_container.sh index aae22cf84..21bdd07a5 100755 --- a/scripts/ci/amd/amd_ci_start_container.sh +++ b/scripts/ci/amd/amd_ci_start_container.sh @@ -23,7 +23,7 @@ fi ROCM_VERSION="rocm700" DEFAULT_MI30X_BASE_TAG="${SGLANG_VERSION}-${ROCM_VERSION}-mi30x" DEFAULT_MI35X_BASE_TAG="${SGLANG_VERSION}-${ROCM_VERSION}-mi35x" -LOCAL_DOCKER_REGISTRY="10.245.143.50:5000" +LOCAL_DOCKER_REGISTRY="10.44.14.109:5000" # Parse command line arguments MI30X_BASE_TAG="${DEFAULT_MI30X_BASE_TAG}" @@ -162,7 +162,7 @@ find_latest_image() { # We intentionally do not probe ${LOCAL_DOCKER_REGISTRY} here with # `docker manifest inspect --insecure` because that command runs in the # runner pod's network namespace, which on every observed AMD scale set - # cannot reach 10.245.143.50:5000 (every probe either fast-fails with TLS + # cannot reach 10.44.14.109:5000 (every probe either fast-fails with TLS # reject or hits a 30s TCP timeout, multiplied across 7 daily candidates). # The actual local-registry pull still happens in the call site below via # `docker pull "${LOCAL_DOCKER_REGISTRY}/${IMAGE}"`, which goes through the diff --git a/scripts/ci/amd/amd_ci_start_container_disagg.sh b/scripts/ci/amd/amd_ci_start_container_disagg.sh index e04d6575e..7d404619b 100755 --- a/scripts/ci/amd/amd_ci_start_container_disagg.sh +++ b/scripts/ci/amd/amd_ci_start_container_disagg.sh @@ -23,7 +23,7 @@ fi ROCM_VERSION="rocm700" DEFAULT_MI30X_BASE_TAG="${SGLANG_VERSION}-${ROCM_VERSION}-mi30x" DEFAULT_MI35X_BASE_TAG="${SGLANG_VERSION}-${ROCM_VERSION}-mi35x" -LOCAL_DOCKER_REGISTRY="10.245.143.50:5000" +LOCAL_DOCKER_REGISTRY="10.44.14.109:5000" # Parse command line arguments MI30X_BASE_TAG="${DEFAULT_MI30X_BASE_TAG}"