Merge branch 'main' into fix/714-build-intelligent-pod-scheduling-based-on-network-latency

Author: OtowoSamuel

.github/CI_COMMANDS.md

@@ -3,7 +3,7 @@
 ## Overview
 
 This document describes the optimized CI/CD pipeline architecture introduced
-to address issues #663, #664, #665, and #666.
+to address issues #700, #701, #703, and #714.
 
 ---
 
@@ -20,7 +20,7 @@ All reusable logic lives under `.github/actions/`:
 
 ---
 
-## Core CI Workflows (#663)
+## Core CI Workflows (#700)
 
 ### `ci.yml`
 - **Change detection** gates expensive jobs (helm-lint, api-docs, examples-smoke-test,
@@ -47,7 +47,7 @@ the critical path by ~35–40% compared to the previous sequential layout.
 
 ---
 
-## Heavy Validation Workflows (#666)
+## Heavy Validation Workflows (#703)
 
 ### `chaos-tests.yml`
 - **Extracted** cluster provisioning into `setup-kind-cluster` composite action.
@@ -70,25 +70,17 @@ the critical path by ~35–40% compared to the previous sequential layout.
 
 ---
 
-## Performance & Benchmark Workflows (#664)
-
-### `benchmark.yml`
-- Build step produces a Docker image saved as a `.tar.gz` artifact.
-- `setup-perf-env` composite action handles k6, kind, kubectl, cluster, RBAC,
-  operator deployment, and port-forwarding — eliminating ~80 lines of duplicated
-  setup across the three benchmark workflows.
-- Baseline comparison uses the same `compare_benchmarks.py` script as
-  `performance-regression.yml`.
-
-### `performance-regression.yml`
-- Removed the now-redundant `setup-cluster` job (consolidated into
-  `performance-test` via `setup-perf-env`).
-- Removed duplicate k6/kind/kubectl install steps.
-
-### `webhook-benchmark.yml`
-- Uses `setup-rust` composite action.
-- Standardised on `actions/upload-artifact@v4` / `actions/download-artifact@v4`
-  (was mixing v7/v8 which don't exist).
+## Performance & Benchmark Workflows (#701)
+
+### `performance.yml` (unified pipeline)
+- **Replaces** the former `benchmark.yml`, `performance-regression.yml`, and
+  `webhook-benchmark.yml` with a single matrix-driven workflow.
+- **Shared build job** produces the operator binary and Docker image once; all
+  three suites (operator, regression, webhook) download the same artifact.
+- **Matrix execution** runs operator and regression suites via `setup-perf-env`,
+  and the webhook suite directly (no kind cluster required).
+- **Shared baseline comparison** via `.github/actions/compare-benchmarks`
+  composite action wrapping `compare_benchmarks.py`.
 
 ---
 

.github/actions/compare-benchmarks/action.yml

@@ -0,0 +1,75 @@
+name: Compare Benchmark Results
+description: >
+  Shared baseline comparison logic used by all performance/benchmark workflows.
+  Wraps benchmarks/scripts/compare_benchmarks.py with consistent inputs/outputs.
+
+inputs:
+  current-file:
+    description: Path to the current benchmark results JSON
+    required: true
+  baseline-file:
+    description: Path to the baseline JSON file
+    required: true
+  threshold:
+    description: Regression threshold percentage
+    required: false
+    default: "10"
+  output-file:
+    description: Path to write the regression report JSON
+    required: false
+    default: results/regression-report.json
+  fail-on-regression:
+    description: Exit with error when regression is detected
+    required: false
+    default: "false"
+
+outputs:
+  overall-passed:
+    description: "true when no regressions exceed the threshold"
+    value: ${{ steps.compare.outputs.overall_passed }}
+
+runs:
+  using: composite
+  steps:
+    - name: Compare with baseline
+      id: compare
+      shell: bash
+      run: |
+        set -euo pipefail
+        CURRENT="${{ inputs.current-file }}"
+        BASELINE="${{ inputs.baseline-file }}"
+        OUTPUT="${{ inputs.output-file }}"
+        THRESHOLD="${{ inputs.threshold }}"
+        FAIL_FLAG=""
+        [[ "${{ inputs.fail-on-regression }}" == "true" ]] && FAIL_FLAG="--fail-on-regression"
+
+        mkdir -p "$(dirname "$OUTPUT")"
+
+        if [[ ! -f "$CURRENT" ]]; then
+          echo "overall_passed=true" >> "$GITHUB_OUTPUT"
+          echo "No current results at $CURRENT — skipping comparison"
+          exit 0
+        fi
+
+        if [[ ! -f "$BASELINE" ]]; then
+          echo "overall_passed=true" >> "$GITHUB_OUTPUT"
+          echo "No baseline at $BASELINE — skipping comparison"
+          exit 0
+        fi
+
+        python3 benchmarks/scripts/compare_benchmarks.py compare \
+          --current "$CURRENT" \
+          --baseline "$BASELINE" \
+          --threshold "$THRESHOLD" \
+          --output "$OUTPUT" \
+          $FAIL_FLAG \
+          --verbose || COMPARE_EXIT=$?
+
+        if [[ -f "$OUTPUT" ]]; then
+          PASSED=$(jq -r '.overall_passed // "true"' "$OUTPUT")
+          echo "overall_passed=$PASSED" >> "$GITHUB_OUTPUT"
+        else
+          echo "overall_passed=true" >> "$GITHUB_OUTPUT"
+        fi
+
+        exit "${COMPARE_EXIT:-0}"

.github/actions/setup-rust/action.yml

@@ -29,9 +29,11 @@ runs:
       run: |
         sudo apt-get update -qq
         sudo apt-get install -y --no-install-recommends \
+          cmake \
+          pkg-config \
+          libssl-dev \
           libsasl2-dev \
-          libcurl4-openssl-dev \
-          pkg-config
+          libcurl4-openssl-dev
 
     - name: Setup Rust cache
       uses: Swatinem/rust-cache@v2