Adding nno dashboard

.github/workflows/generate_matrix_page.yaml

@@ -32,8 +32,12 @@ jobs:
     steps:
       - name: Set dynamic env vars
         run: |
+          # GPU Operator dashboard paths
           echo "DASHBOARD_DATA_FILEPATH=${DASHBOARD_OUTPUT_DIR}/gpu_operator_matrix.json" >> "$GITHUB_ENV"
           echo "DASHBOARD_HTML_FILEPATH=${DASHBOARD_OUTPUT_DIR}/gpu_operator_matrix.html" >> "$GITHUB_ENV"
+          # Network Operator dashboard paths
+          echo "NNO_DASHBOARD_DATA_FILEPATH=${DASHBOARD_OUTPUT_DIR}/network_operator_matrix.json" >> "$GITHUB_ENV"
+          echo "NNO_DASHBOARD_HTML_FILEPATH=${DASHBOARD_OUTPUT_DIR}/network_operator_matrix.html" >> "$GITHUB_ENV"
           echo "GH_PAGES_BRANCH=${{ github.event.inputs.gh_pages_branch || 'gh-pages' }}" >> "$GITHUB_ENV"
         env:
           DASHBOARD_OUTPUT_DIR: ${{ env.DASHBOARD_OUTPUT_DIR }}
@@ -67,27 +71,62 @@ jobs:
       - name: Install Dependencies
         run: |
           pip install -r workflows/gpu_operator_dashboard/requirements.txt
+          pip install -r workflows/nno_dashboard/requirements.txt
 
       - name: Fetch CI Data
         run: |
           echo "Processing PR: ${{ steps.determine_pr.outputs.PR_NUMBER }}"
+          # GPU Operator
           python -m workflows.gpu_operator_dashboard.fetch_ci_data \
             --pr_number "${{ steps.determine_pr.outputs.PR_NUMBER }}" \
             --baseline_data_filepath "${{ env.DASHBOARD_DATA_FILEPATH }}" \
             --merged_data_filepath "${{ env.DASHBOARD_DATA_FILEPATH }}"
+          # Network Operator
+          python -m workflows.nno_dashboard.fetch_ci_data \
+            --pr_number "${{ steps.determine_pr.outputs.PR_NUMBER }}" \
+            --baseline_data_filepath "${{ env.NNO_DASHBOARD_DATA_FILEPATH }}" \
+            --merged_data_filepath "${{ env.NNO_DASHBOARD_DATA_FILEPATH }}"
 
 
       - name: Generate HTML Dashboard (only if JSON changed)
         run: |
           cd "${{ env.DASHBOARD_OUTPUT_DIR }}"
+
+          # Check if GPU Operator JSON changed
+          GPU_CHANGED=false
           if [[ ${{ github.event_name }} == "pull_request_target" ]] && git diff --exit-code gpu_operator_matrix.json; then
-            echo "no changes"
+            echo "GPU Operator: no changes"
+          else
+            echo "GPU Operator: changes detected"
+            GPU_CHANGED=true
+          fi
+
+          # Check if Network Operator JSON changed
+          NNO_CHANGED=false
+          if [[ ${{ github.event_name }} == "pull_request_target" ]] && git diff --exit-code network_operator_matrix.json; then
+            echo "Network Operator: no changes"
           else
-            cd "${{ github.workspace }}"
+            echo "Network Operator: changes detected"
+            NNO_CHANGED=true
+          fi
+
+          cd "${{ github.workspace }}"
+
+          # Generate GPU Operator dashboard if changed
+          if [ "$GPU_CHANGED" = true ]; then
+            echo "Generating GPU Operator dashboard..."
             python -m workflows.gpu_operator_dashboard.generate_ci_dashboard \
               --dashboard_data_filepath "${{ env.DASHBOARD_DATA_FILEPATH }}" \
               --dashboard_html_filepath "${{ env.DASHBOARD_HTML_FILEPATH }}"
           fi
+
+          # Generate Network Operator dashboard if changed
+          if [ "$NNO_CHANGED" = true ]; then
+            echo "Generating Network Operator dashboard..."
+            python -m workflows.nno_dashboard.generate_ci_dashboard \
+              --dashboard_data_filepath "${{ env.NNO_DASHBOARD_DATA_FILEPATH }}" \
+              --dashboard_html_filepath "${{ env.NNO_DASHBOARD_HTML_FILEPATH }}"
+          fi
 
       - name: Deploy HTML to GitHub Pages
         uses: JamesIves/github-pages-deploy-action@v4

workflows/gpu_operator_dashboard/fetch_ci_data.py

@@ -92,15 +92,20 @@ class TestResult:
     test_status: str
     prow_job_url: str
     job_timestamp: str
+    test_flavor: Optional[str] = None  # NNO-specific: test configuration flavor
 
     def to_dict(self) -> Dict[str, Any]:
-        return {
+        result = {
             OCP_FULL_VERSION: self.ocp_full_version,
             GPU_OPERATOR_VERSION: self.gpu_operator_version,
             "test_status": self.test_status,
             "prow_job_url": self.prow_job_url,
             "job_timestamp": self.job_timestamp,
         }
+        # Include test_flavor only if it's set (NNO-specific)
+        if self.test_flavor is not None:
+            result["test_flavor"] = self.test_flavor
+        return result
 
     def composite_key(self) -> TestResultKey:
         repo, pr_number, job_name, build_id = extract_build_components(self.prow_job_url)
@@ -571,8 +576,15 @@ def merge_ocp_version_results(
     bundle_result_limit: Optional[int] = None
 ) -> Dict[str, Any]:
     """Merge results for a single OCP version."""
-    # Initialize the structure
-    merged_version_data = {"notes": [], "bundle_tests": [], "release_tests": [], "job_history_links": []}
+    # Initialize the structure with all possible fields
+    merged_version_data = {
+        "notes": [],
+        "bundle_tests": [],
+        "release_tests": [],
+        "job_history_links": [],
+        "test_flavors": {}
+    }
+    # Update with existing data (preserves any additional fields)
     merged_version_data.update(existing_version_data)
 
     # Merge bundle tests with limit
@@ -599,6 +611,31 @@ def merge_ocp_version_results(
     # Convert back to sorted list for JSON serialization
     merged_version_data["job_history_links"] = sorted(list(all_job_history_links))
 
+    # Merge test_flavors (NNO-specific) if present
+    new_test_flavors = new_version_data.get("test_flavors", {})
+    existing_test_flavors = merged_version_data.get("test_flavors", {})
+
+    # Merge test flavors by combining results for each flavor
+    for flavor_name, flavor_data in new_test_flavors.items():
+        if flavor_name not in existing_test_flavors:
+            existing_test_flavors[flavor_name] = {"results": [], "job_history_links": set()}
+
+        # Merge results for this flavor (using same logic as release_tests)
+        new_flavor_results = flavor_data.get("results", [])
+        existing_flavor_results = existing_test_flavors[flavor_name].get("results", [])
+        existing_test_flavors[flavor_name]["results"] = merge_release_tests(
+            new_flavor_results, existing_flavor_results
+        )
+
+        # Merge job history links for this flavor
+        new_flavor_links = flavor_data.get("job_history_links", set())
+        existing_flavor_links = existing_test_flavors[flavor_name].get("job_history_links", set())
+        all_flavor_links = set(existing_flavor_links if isinstance(existing_flavor_links, (set, list)) else [])
+        all_flavor_links.update(new_flavor_links)
+        existing_test_flavors[flavor_name]["job_history_links"] = sorted(list(all_flavor_links))
+
+    merged_version_data["test_flavors"] = existing_test_flavors
+
     return merged_version_data
 
 

workflows/nno_dashboard/README.md

@@ -0,0 +1,136 @@
+# NVIDIA Network Operator Dashboard Workflow
+
+This workflow generates an HTML dashboard showing NVIDIA Network Operator test results across different operator versions and OpenShift versions. It fetches test data from CI systems and creates visual reports for tracking test status over time.
+
+## Overview
+
+The dashboard workflow:
+- Fetches test results from Google Cloud Storage based on pull request data
+- Supports various network operator test patterns including:
+  - `nvidia-network-operator-legacy-sriov-rdma`
+  - `nvidia-network-operator-e2e`
+  - DOCA-based tests (e.g., `doca4-nvidia-network-operator-*`)
+- Merges new results with existing baseline data
+- Generates HTML dashboard reports
+- Automatically deploys updates to GitHub Pages
+
+## Architecture
+
+This dashboard **reuses** the GPU Operator Dashboard code and only overrides the operator-specific parts:
+- ✅ Imports all core logic from `workflows.gpu_operator_dashboard.fetch_ci_data`
+- ✅ Overrides only Network Operator specific:
+  - Regex patterns to match network operator job names
+  - Artifact paths (`network-operator-e2e/artifacts/`)
+  - Version field names (`network_operator_version` vs `gpu_operator_version`)
+- ✅ Maintains a clean, DRY codebase with minimal duplication
+
+This design makes maintenance easier - bug fixes in the core logic automatically benefit both dashboards.
+
+## Supported Test Patterns
+
+The dashboard recognizes the following test job patterns:
+- `pull-ci-rh-ecosystem-edge-nvidia-ci-main-{version}-nvidia-network-operator-legacy-sriov-rdma`
+- `pull-ci-rh-ecosystem-edge-nvidia-ci-main-{version}-nvidia-network-operator-e2e`
+- `rehearse-{id}-pull-ci-rh-ecosystem-edge-nvidia-ci-main-doca4-nvidia-network-operator-*`
+
+Example URL that will be processed:
+```
+https://gcsweb-ci.apps.ci.l2s4.p1.openshiftapps.com/gcs/test-platform-results/pr-logs/pull/openshift_release/67673/rehearse-67673-pull-ci-rh-ecosystem-edge-nvidia-ci-main-doca4-nvidia-network-operator-legacy-sriov-rdma/1961127149603655680/
+```
+
+## Usage
+
+### Prerequisites
+
+```console
+pip install -r workflows/nno_dashboard/requirements.txt
+```
+
+**Important:** Before running fetch_ci_data.py, create the baseline data file and initialize it with an empty JSON object if it doesn't exist:
+
+```console
+echo '{}' > nno_data.json
+```
+
+### Fetch CI Data
+
+```console
+# Process a specific PR
+python -m workflows.nno_dashboard.fetch_ci_data --pr_number "123" --baseline_data_filepath nno_data.json --merged_data_filepath nno_data.json
+
+# Process all merged PRs - limited to 100 most recent (default)
+python -m workflows.nno_dashboard.fetch_ci_data --pr_number "all" --baseline_data_filepath nno_data.json --merged_data_filepath nno_data.json
+
+# Process with bundle result limit (keep only last 50 bundle tests per version)
+python -m workflows.nno_dashboard.fetch_ci_data --pr_number "all" --baseline_data_filepath nno_data.json --merged_data_filepath nno_data.json --bundle_result_limit 50
+```
+
+### Generate Dashboard
+
+```console
+python -m workflows.nno_dashboard.generate_ci_dashboard --dashboard_data_filepath nno_data.json --dashboard_html_filepath nno_dashboard.html
+```
+
+The dashboard generator also **reuses** the GPU Operator dashboard code:
+- Imports all HTML generation logic from `workflows.gpu_operator_dashboard.generate_ci_dashboard`
+- Uses Network Operator specific templates (in `templates/` directory)
+- Only aliases `NETWORK_OPERATOR_VERSION` as `GPU_OPERATOR_VERSION` for compatibility
+
+### Running Tests
+
+First, make sure `pytest` is installed. Then, run:
+
+```console
+python -m pytest workflows/nno_dashboard/tests/ -v
+```
+
+## GitHub Actions Integration
+
+- **Automatic**: Processes merged pull requests to update the dashboard with new test results and deploys to GitHub Pages
+- **Manual**: Can be triggered manually via GitHub Actions workflow dispatch
+
+## Data Structure
+
+The fetched data follows this structure:
+
+```json
+{
+  "doca4": {
+    "notes": [],
+    "bundle_tests": [
+      {
+        "ocp_full_version": "4.16.0",
+        "network_operator_version": "24.10.0",
+        "test_status": "SUCCESS",
+        "prow_job_url": "https://...",
+        "job_timestamp": "1234567890"
+      }
+    ],
+    "release_tests": [...],
+    "job_history_links": [
+      "https://prow.ci.openshift.org/job-history/gs/test-platform-results/pr-logs/directory/..."
+    ]
+  }
+}
+```
+
+## Troubleshooting
+
+### No data being fetched
+
+1. Verify the PR number exists and has network operator test runs
+2. Check that the job names match the expected patterns (see regex in fetch_ci_data.py line 36-40)
+3. Ensure the test artifacts contain the required files:
+   - `finished.json`
+   - `network-operator-e2e/artifacts/ocp.version`
+   - `network-operator-e2e/artifacts/operator.version`
+
+### Regex pattern not matching
+
+The regex pattern is designed to match:
+- Repository: `rh-ecosystem-edge_nvidia-ci` or `openshift_release` (for rehearse jobs)
+- OCP version prefix: Can be `doca4`, `nno1`, or other custom prefixes
+- Job suffix: Must contain `nvidia-network-operator` followed by test type
+
+If your job names don't match, you may need to adjust the `TEST_RESULT_PATH_REGEX` pattern in `fetch_ci_data.py`.
+