From 9f9acf0405ad5fb623def9b76439b8d5b55177e1 Mon Sep 17 00:00:00 2001
From: Siarhei Fedartsou <siarhei.fedartsou@gmail.com>
Date: Sun, 26 May 2024 15:01:39 +0200
Subject: [PATCH] Implement end-to-end benchmark

---
 .github/workflows/osrm-backend.yml |  3 +-
 scripts/ci/download_gps_traces.py  | 91 ++++++++++++++++++++++++++++++
 scripts/ci/run_benchmarks.sh       | 12 +++-
 3 files changed, 103 insertions(+), 3 deletions(-)
 create mode 100644 scripts/ci/download_gps_traces.py

diff --git a/.github/workflows/osrm-backend.yml b/.github/workflows/osrm-backend.yml
index 4943d6de3..83ea2587d 100644
--- a/.github/workflows/osrm-backend.yml
+++ b/.github/workflows/osrm-backend.yml
@@ -641,8 +641,9 @@ jobs:
           restore-keys: |
             v1-conan-benchmarks-
       - name: Prepare data
-        run: | 
+        run: |
           wget http://download.geofabrik.de/europe/germany/berlin-latest.osm.pbf -O ~/data.osm.pbf
+          python3 scripts/ci/download_gps_traces.py 13.0884 52.3383 13.7611 52.6755 > ~/gps_traces.csv
       - name: Install dependencies
         run: python3 -m pip install "conan<2.0.0" "requests==2.31.0" "locust==2.28.0"
       - name: Checkout PR Branch
diff --git a/scripts/ci/download_gps_traces.py b/scripts/ci/download_gps_traces.py
new file mode 100644
index 000000000..961acd532
--- /dev/null
+++ b/scripts/ci/download_gps_traces.py
@@ -0,0 +1,91 @@
+import requests
+import xml.etree.ElementTree as ET
+import csv
+import sys
+import argparse
+
+def get_osm_gps_traces(min_lon, min_lat, max_lon, max_lat):
+    url = 'https://api.openstreetmap.org/api/0.6/trackpoints'
+    traces = []
+    
+    lon_step = 0.25
+    lat_step = 0.25
+    
+    current_min_lon = min_lon
+
+    while current_min_lon < max_lon:
+        current_max_lon = min(current_min_lon + lon_step, max_lon)
+        
+        current_min_lat = min_lat
+        while current_min_lat < max_lat:
+            current_max_lat = min(current_min_lat + lat_step, max_lat)
+            
+            bbox = f'{current_min_lon},{current_min_lat},{current_max_lon},{current_max_lat}'
+            print(f"Requesting bbox: {bbox}", file=sys.stderr)
+            
+            params = {
+                'bbox': bbox,
+                'page': 0
+            }
+            headers = {
+                'Accept': 'application/xml'
+            }
+            
+            response = requests.get(url, params=params, headers=headers)
+            if response.status_code == 200:
+                traces.append(response.content)
+            else:
+                print(f"Error fetching data for bbox {bbox}: {response.status_code} {response.text}", file=sys.stderr)
+            
+            current_min_lat += lat_step
+        current_min_lon += lon_step
+    
+    return traces
+
+def parse_gpx_data(gpx_data):
+    try:
+        root = ET.fromstring(gpx_data)
+    except ET.ParseError as e:
+        print(f"Error parsing GPX data: {e}", file=sys.stderr)
+        return []
+    namespace = {'gpx': 'http://www.topografix.com/GPX/1/0'}
+
+    tracks = []
+    for trk in root.findall('.//gpx:trk', namespace):
+        track_data = []
+        for trkseg in trk.findall('.//gpx:trkseg', namespace):
+            for trkpt in trkseg.findall('gpx:trkpt', namespace):
+                lat = trkpt.get('lat')
+                lon = trkpt.get('lon')
+                time = trkpt.find('time').text if trkpt.find('time') is not None else ''
+                track_data.append([lat, lon, time])
+        tracks.append(track_data)
+    return tracks
+
+def save_to_csv(data, file):
+    writer = csv.writer(file)
+    writer.writerow(['TrackID', 'Latitude', 'Longitude', 'Time'])
+    writer.writerows(data)
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser(description='Fetch and output OSM GPS traces for a given bounding box.')
+    parser.add_argument('min_lon', type=float, help='Minimum longitude of the bounding box')
+    parser.add_argument('min_lat', type=float, help='Minimum latitude of the bounding box')
+    parser.add_argument('max_lon', type=float, help='Maximum longitude of the bounding box')
+    parser.add_argument('max_lat', type=float, help='Maximum latitude of the bounding box')
+    
+    args = parser.parse_args()
+    
+    gpx_data_traces = get_osm_gps_traces(args.min_lon, args.min_lat, args.max_lon, args.max_lat)
+    print(f"Collected {len(gpx_data_traces)} trace segments", file=sys.stderr)
+
+    all_data = []
+    track_id = 0
+    for gpx_data in gpx_data_traces:
+        for track in parse_gpx_data(gpx_data):
+            for point in track:
+                all_data.append([track_id] + point)
+            track_id += 1
+    
+    # Output all data to stdout
+    save_to_csv(all_data, sys.stdout)
diff --git a/scripts/ci/run_benchmarks.sh b/scripts/ci/run_benchmarks.sh
index 1cdd76a66..c551cb998 100755
--- a/scripts/ci/run_benchmarks.sh
+++ b/scripts/ci/run_benchmarks.sh
@@ -31,9 +31,17 @@ function run_benchmarks_for_folder {
 
     # TODO: save results
     curl --retry-delay 3 --retry 10 --retry-all-errors "http://127.0.0.1:5000/route/v1/driving/13.388860,52.517037;13.385983,52.496891?steps=true"
-    locust -f locustfile.py --headless --users 10 --spawn-rate 1 --host http://localhost:5000 --run-time 1m --csv=results
+    locust -f $FOLDER/locustfile.py --headless --users 10 --spawn-rate 1 --host http://localhost:5000 --run-time 1m --csv=results
+
+    echo "STATS: "
+    cat results_stats.csv
+    echo "FAILURES: "
+    cat results_failures.csv
+    echo "EXCEPTIONS: "
+    cat results_exceptions.csv
+    echo "STATS HISTORY: "
+    cat results_stats_history.csv
 
-    cat results_distribution.csv
 
     kill -0 $OSRM_ROUTED_PID
 }