[BE] Use `teardown-linux`/`chown` actions for binary builds (#84449)
Also embed `wait_for_ssh_to_drain.sh` into the action (to make it more reusable across repos) and delete unused teardown_linux template from `common.yml`
Pull Request resolved: https://github.com/pytorch/pytorch/pull/84449
Approved by: https://github.com/kit1980
diff --git a/.github/actions/teardown-linux/action.yml b/.github/actions/teardown-linux/action.yml
index 9238a07..024bb3c 100644
--- a/.github/actions/teardown-linux/action.yml
+++ b/.github/actions/teardown-linux/action.yml
@@ -16,7 +16,18 @@
# Always hold for active ssh sessions
shell: bash
if: inputs.skip-wait-ssh == ''
- run: .github/scripts/wait_for_ssh_to_drain.sh
+ run: |
+ set -eou pipefail
+
+ echo "Holding runner for 2 hours until all ssh sessions have logged out"
+ for _ in $(seq 1440); do
+ # Break if no ssh session exists anymore
+ if [ "$(who)" = "" ]; then
+ break
+ fi
+ echo "."
+ sleep 5
+ done
- name: Kill containers, clean up images
shell: bash
diff --git a/.github/scripts/wait_for_ssh_to_drain.sh b/.github/scripts/wait_for_ssh_to_drain.sh
deleted file mode 100755
index f33d807..0000000
--- a/.github/scripts/wait_for_ssh_to_drain.sh
+++ /dev/null
@@ -1,13 +0,0 @@
-#!/usr/bin/env bash
-
-set -eou pipefail
-
-echo "Holding runner for 2 hours until all ssh sessions have logged out"
-for _ in $(seq 1440); do
- # Break if no ssh session exists anymore
- if [ "$(who)" = "" ]; then
- break
- fi
- echo "."
- sleep 5
-done
diff --git a/.github/templates/common.yml.j2 b/.github/templates/common.yml.j2
index b80b82f..37f89ef 100644
--- a/.github/templates/common.yml.j2
+++ b/.github/templates/common.yml.j2
@@ -199,29 +199,6 @@
env | grep '^CI' >> "/tmp/github_env_${GITHUB_RUN_ID}"
{%- endmacro -%}
-{%- macro teardown_ec2_linux(pytorch_directory="") -%}
- - name: Hold runner for 2 hours or until ssh sessions have drained
-{%- if pytorch_directory %}
- working-directory: !{{ pytorch_directory }}
-{%- endif %}
- # Always hold for active ssh sessions
- if: always()
- run: .github/scripts/wait_for_ssh_to_drain.sh
- - name: Chown workspace
- if: always()
- run: |
- # Ensure the working directory gets chowned back to the current user
- docker run --rm -v "$(pwd)":/v -w /v "${ALPINE_IMAGE}" chown -R "$(id -u):$(id -g)" .
- - name: Kill containers, clean up images
- if: always()
- run: |
- # ignore expansion of "docker ps -q" since it could be empty
- # shellcheck disable=SC2046
- docker stop $(docker ps -q) || true
- # Prune all of the docker images
- docker system prune -af
-{%- endmacro -%}
-
{%- macro teardown_rocm_linux() -%}
- name: Kill containers, clean up images
if: always()
diff --git a/.github/workflows/_binary-build-linux.yml b/.github/workflows/_binary-build-linux.yml
index dc69e3a..8ddd4af 100644
--- a/.github/workflows/_binary-build-linux.yml
+++ b/.github/workflows/_binary-build-linux.yml
@@ -167,11 +167,9 @@
run: |
echo "BUILD_SPLIT_CUDA='ON'" >> "$GITHUB_ENV"
- name: Pull Docker image
- run: |
- retry () {
- "$@" || (sleep 1 && "$@") || (sleep 2 && "$@")
- }
- retry docker pull "${DOCKER_IMAGE}"
+ uses: ./pytorch/.github/actions/pull-docker-image
+ with:
+ docker-image: ${{ inputs.DOCKER_IMAGE }}
- name: Build PyTorch binary
run: |
set -x
@@ -204,10 +202,7 @@
docker exec -t "${container_name}" bash -c "source ${BINARY_ENV_FILE} && bash /builder/${{ inputs.PACKAGE_TYPE }}/build.sh"
- name: Chown artifacts
if: always()
- shell: bash
- run: |
- # Ensure the working directory gets chowned back to the current user
- docker run --rm -v "${RUNNER_TEMP}/artifacts:/v" -w /v "${ALPINE_IMAGE}" chown -R "$(id -u):$(id -g)" .
+ uses: ./pytorch/.github/actions/chown-workspace
- uses: actions/upload-artifact@v3
with:
@@ -216,21 +211,10 @@
path:
${{ runner.temp }}/artifacts/*
- - name: Hold runner for 2 hours or until ssh sessions have drained
- working-directory: pytorch/
- # Always hold for active ssh sessions
+ - name: Teardown Linux
if: always()
- run: .github/scripts/wait_for_ssh_to_drain.sh
+ uses: ./pytorch/.github/actions/teardown-linux
+
- name: Chown workspace
if: always()
- run: |
- # Ensure the working directory gets chowned back to the current user
- docker run --rm -v "$(pwd)":/v -w /v "${ALPINE_IMAGE}" chown -R "$(id -u):$(id -g)" .
- - name: Kill containers, clean up images
- if: always()
- run: |
- # ignore expansion of "docker ps -q" since it could be empty
- # shellcheck disable=SC2046
- docker stop $(docker ps -q) || true
- # Prune all of the docker images
- docker system prune -af
+ uses: ./pytorch/.github/actions/chown-workspace
diff --git a/.github/workflows/_binary-test-linux.yml b/.github/workflows/_binary-test-linux.yml
index e8749c5..c5e9fab 100644
--- a/.github/workflows/_binary-test-linux.yml
+++ b/.github/workflows/_binary-test-linux.yml
@@ -192,21 +192,10 @@
- name: Test Pytorch binary
uses: ./pytorch/.github/actions/test-pytorch-binary
- - name: Hold runner for 2 hours or until ssh sessions have drained
- working-directory: pytorch/
- # Always hold for active ssh sessions
+ - name: Teardown Linux
if: always()
- run: .github/scripts/wait_for_ssh_to_drain.sh
+ uses: ./pytorch/.github/actions/teardown-linux
+
- name: Chown workspace
if: always()
- run: |
- # Ensure the working directory gets chowned back to the current user
- docker run --rm -v "$(pwd)":/v -w /v "${ALPINE_IMAGE}" chown -R "$(id -u):$(id -g)" .
- - name: Kill containers, clean up images
- if: always()
- run: |
- # ignore expansion of "docker ps -q" since it could be empty
- # shellcheck disable=SC2046
- docker stop $(docker ps -q) || true
- # Prune all of the docker images
- docker system prune -af
+ uses: ./pytorch/.github/actions/chown-workspace
