paper data: utils iperf.py

results/paper_data/utils/iperf.py

+import glob
+import json
+import os
+import fnmatch
+import re
+import itertools
+import sys
+
+def parse_iperf_run(data, skip=1, use=8):
+    tp_pat = re.compile(r'\[ *\d*\] *([0-9\.]*)- *([0-9\.]*) sec.*Bytes *([0-9\.]*) ([GM])bits.*')
+    tps_time = {}
+    for hn in fnmatch.filter(data['sims'].keys(), 'host.client.*'):
+        sim = data['sims'][hn]
+        for l in sim['stdout']:
+            m = tp_pat.match(l)
+            if not m:
+                continue
+
+            time = int(float(m.group(1)))
+            if time < skip:
+                continue
+            if time >= skip + use:
+                continue
+
+            if not time in tps_time:
+                tps_time[time] = []
+
+            if m.group(4) == 'G':
+                tps_time[time].append(float(m.group(3)))
+            elif m.group(4) == 'M':
+                m_tps = float(m.group(3))/1000
+                tps_time[time].append(m_tps)
+
+    tps = []
+    for t in sorted(tps_time.keys()):
+        x = sum(tps_time[t])
+        tps.append(x)
+
+
+    if len(tps) == 0:
+        return None
+    return sum(tps) / len(tps)
+
+def parse_iperf(basename, skip=1, use=8):
+    runs = []
+    for path in glob.glob(basename + '-*.json'):
+        if path == basename + '-0.json':
+            # skip checkpoints
+            continue
+
+        with open(path, 'r') as f:
+            data = json.load(f)
+        result = parse_iperf_run(data, skip, use)
+        if result is not None:
+            runs.append(result)
+
+    if not runs:
+        return {'avg': None, 'min': None, 'max': None}
+    else:
+        return {'avg': sum(runs) / len(runs), 'min': min(runs),
+                'max': max(runs)}
+    result = {}