Add CI slack notifications.

.github/actions/workflow-build/action.yml

@@ -25,6 +25,15 @@ inputs:
     description: "Path to the matrix parser script (default if blank: build-workflow.py from action dir)"
     default: ""
     required: false
+  slack_token:
+    description: "The Slack token to use for notifications. No notifications will be sent if not provided."
+    required: false
+  slack_log:
+    description: "Slack channel ID for verbose notifications."
+    required: false
+  slack_alert:
+    description: "Slack channel ID for alert notifications."
+    required: false
 
 outputs:
   workflow:
@@ -35,6 +44,20 @@ runs:
   using: "composite"
   steps:
 
+    - name: Send Slack log notification
+      if: ${{inputs.slack_token != '' && inputs.slack_log != '' }}
+      uses: slackapi/[email protected]
+      env:
+        SLACK_BOT_TOKEN: ${{ inputs.slack_token }}
+        WORKFLOW_TYPE: ${{ github.workflow }} # nightly, weekly, pr, etc.
+        SUMMARY_URL: https://github.com/${{github.repository}}/actions/runs/${{github.run_id}}
+      with:
+        channel-id: ${{ inputs.slack_log }}
+        slack-message: |
+          Workflow '${{ env.WORKFLOW_TYPE }}' starting...
+
+          Details: ${{ env.SUMMARY_URL }}
+
     - name: Inspect changes
       if: ${{ inputs.inspect_changes_script != '' && inputs.inspect_changes_base_sha != '' }}
       id: inspect-changes
@@ -99,3 +122,18 @@ runs:
         name: workflow
         path: workflow/
         compression-level: 0
+
+    - name: Send Slack error notification
+      if: ${{ failure() && inputs.slack_token != '' && (inputs.slack_alert != '' || inputs.slack_log != '') }}
+      uses: slackapi/[email protected]
+      env:
+        SLACK_BOT_TOKEN: ${{ inputs.slack_token }}
+        WORKFLOW_TYPE: ${{ github.workflow }} # nightly, weekly, pr, etc.
+        SUMMARY_URL: https://github.com/${{github.repository}}/actions/runs/${{github.run_id}}
+        CHANNEL_SEP: ${{ (inputs.slack_log != '' && inputs.slack_alert != '') && ',' || ''}}
+      with:
+        channel-id: '${{ inputs.slack_log }}${{env.CHANNEL_SEP}}${{ inputs.slack_alert }}'
+        slack-message: |
+          Workflow '${{ env.WORKFLOW_TYPE }}' encountered an error while preparing to run.
+
+          Details: ${{ env.SUMMARY_URL }}

.github/actions/workflow-results/action.yml

@@ -8,11 +8,20 @@ inputs:
   pr_number:
     description: "The PR number to comment on, if applicable. No comment will be made if not provided."
     required: false
+  slack_token:
+    description: "The Slack token to use for notifications. No notifications will be sent if not provided."
+    required: false
+  slack_log:
+    description: "Slack channel ID for verbose notifications."
+    required: false
+  slack_alert:
+    description: "Slack channel ID for alert notifications."
+    required: false
 
 outputs:
   success:
     description: "Whether any jobs failed."
-    value: ${{ steps.check-dispatch.outputs.success }}
+    value: ${{ steps.check-success.outputs.success }}
 
 runs:
   using: "composite"
@@ -112,6 +121,10 @@ runs:
         printf "SUMMARY=%s\n" "$(cat final_summary.md | url_encode_string)" | tee -a "${GITHUB_OUTPUT}"
         echo "::endgroup::"
 
+        echo "::group::GHA Output: EXEC_SUMMARY"
+        printf "EXEC_SUMMARY=%s\n" "$(cat execution/heading.txt)" | tee -a "${GITHUB_OUTPUT}"
+        echo "::endgroup::"
+
         cp final_summary.md ${GITHUB_STEP_SUMMARY}
 
     - name: Comment on PR
@@ -140,7 +153,7 @@ runs:
           });
 
     - name: Check for job success
-      id: check-dispatch
+      id: check-success
       shell: bash --noprofile --norc -euo pipefail {0}
       run: |
         echo "::group::Checking for success artifacts"
@@ -162,3 +175,38 @@ runs:
         fi
 
         echo "success=true" >> "${GITHUB_OUTPUT}"
+
+    - name: Send Slack log notification
+      if: ${{ always() && inputs.slack_token != '' && inputs.slack_log != '' }}
+      uses: slackapi/[email protected]
+      env:
+        SLACK_BOT_TOKEN: ${{ inputs.slack_token }}
+        WORKFLOW_TYPE: ${{ github.workflow }} # nightly, weekly, pr, etc.
+        STATUS: ${{ steps.check-success.outcome }}
+        EXEC_SUMMARY: ${{ steps.final-summary.outputs.EXEC_SUMMARY }}
+        SUMMARY_URL: https://github.com/${{github.repository}}/actions/runs/${{github.run_id}}
+      with:
+        channel-id: ${{ inputs.slack_log }}
+        slack-message: |
+          Workflow '${{ env.WORKFLOW_TYPE }}' has finished with status `${{ env.STATUS }}`:
+
+          ${{ env.EXEC_SUMMARY }}
+
+          Details: ${{ env.SUMMARY_URL }}
+
+    - name: Send Slack alert notification
+      if: ${{ failure() && inputs.slack_token != '' && inputs.slack_alert != '' }}
+      uses: slackapi/[email protected]
+      env:
+        SLACK_BOT_TOKEN: ${{ inputs.slack_token }}
+        WORKFLOW_TYPE: ${{ github.workflow }} # nightly, weekly, pr, etc.
+        EXEC_SUMMARY: ${{ steps.final-summary.outputs.EXEC_SUMMARY }}
+        SUMMARY_URL: https://github.com/${{github.repository}}/actions/runs/${{github.run_id}}
+      with:
+        channel-id: ${{ inputs.slack_alert }}
+        slack-message: |
+          Workflow '${{ env.WORKFLOW_TYPE }}' has failed:
+
+          ${{ env.EXEC_SUMMARY }}
+
+          Details: ${{ env.SUMMARY_URL }}

.github/actions/workflow-results/parse-job-times.py

@@ -120,7 +120,7 @@ def main():
     for id, stats in result.items():
         job_seconds = stats['job_seconds']
         command_seconds = stats['command_seconds']
-        overhead = (job_seconds - command_seconds) * 100 / command_seconds
+        overhead = (job_seconds - command_seconds) * 100 / command_seconds if command_seconds > 0 else 100
         print(f"{stats['job_duration']:10} {stats['command_duration']:10} {overhead:10.0f} {stats['name']}")
     print("::endgroup::")
 

.github/workflows/build-rapids.yml

@@ -2,6 +2,15 @@ name: Build all RAPIDS repositories
 
 on:
   workflow_call:
+    inputs:
+      slack_token:
+        description: "The Slack token to use for notifications. No notifications will be sent if not provided."
+        required: false
+        type: string
+      slack_alert:
+        description: "Slack channel ID for alert notifications."
+        required: false
+        type: string
 
 jobs:
   check-event:
@@ -156,3 +165,22 @@ jobs:
             --volume "$RUNNER_TEMP/ci.sh:/ci.sh" \
             --volume "$RUNNER_TEMP/ci-entrypoint.sh:/ci-entrypoint.sh" \
             -- /ci-entrypoint.sh ./ci/rapids/rapids-entrypoint.sh /ci.sh
+
+  notify-failure:
+    name: Notify Slack of RAPIDS failure
+    if: ${{ failure() && inputs.slack_token != '' && inputs.slack_alert != '' }}
+    needs: build-rapids
+    runs-on: ubuntu-latest
+    steps:
+      - name: Notify
+        uses: slackapi/[email protected]
+        env:
+          SLACK_BOT_TOKEN: ${{ inputs.slack_token }}
+          WORKFLOW_TYPE: ${{ github.workflow }}
+          SUMMARY_URL: https://github.com/${{github.repository}}/actions/runs/${{github.run_id}}
+        with:
+          channel-id: ${{ inputs.slack_alert }}
+          slack-message: |
+            RAPIDS build in workflow '${{ env.WORKFLOW_TYPE }}' failed.
+
+            Details: ${{ env.SUMMARY_URL }}

.github/workflows/ci-workflow-nightly.yml

@@ -46,6 +46,9 @@ jobs:
         uses: ./.github/actions/workflow-build
         with:
           workflows: nightly
+          slack_token: ${{ secrets.SLACK_NOTIFIER_BOT_TOKEN }}
+          slack_log: ${{ secrets.SLACK_CHANNEL_CI_LOG }}
+          slack_alert: ${{ secrets.SLACK_CHANNEL_CI_ALERT }}
 
   dispatch-groups-linux-two-stage:
     name: ${{ matrix.name }}
@@ -128,6 +131,11 @@ jobs:
       - name: Check workflow success
         id: check-workflow
         uses: ./.github/actions/workflow-results
+        with:
+          slack_token: ${{ secrets.SLACK_NOTIFIER_BOT_TOKEN }}
+          slack_log: ${{ secrets.SLACK_CHANNEL_CI_LOG }}
+          slack_alert: ${{ secrets.SLACK_CHANNEL_CI_ALERT }}
+
 
   build-rapids:
     name: Build RAPIDS
@@ -139,38 +147,6 @@ jobs:
       contents: read
       pull-requests: read
     uses: ./.github/workflows/build-rapids.yml
-
-  # Check all other job statuses. This job gates branch protection checks.
-  ci:
-    name: CI
-    # !! Important: This job is used for branch protection checks.
-    # !! Need to use always() instead of !cancelled() because skipped jobs count as success
-    # !! for Github branch protection checks. Yes, really: by default, branch protections
-    # !! can be bypassed by cancelling CI. See NVIDIA/cccl#605.
-    if: ${{ always() }}
-    needs:
-      - verify-workflow
-    runs-on: ubuntu-latest
-    steps:
-      - name: Check results
-        run: |
-          status="success"
-
-          check_result() {
-            name=$1
-            expected=$2
-            result=$3
-
-            echo "Checking if $name job result ('$result') is '$expected'..."
-            if [[ "$result" != "$expected" ]]; then
-              echo "$name job failed"
-
-              status="failed"
-            fi
-          }
-
-          check_result "verify-workflow"      "success" "${{needs.verify-workflow.result}}"
-
-          if [[ "$status" != "success" ]]; then
-            exit 1
-          fi
+    with:
+      slack_token: ${{ secrets.SLACK_NOTIFIER_BOT_TOKEN }}
+      slack_alert: ${{ secrets.SLACK_CHANNEL_CI_ALERT }}