From 4b7f1752eca525f1893ec3fe540b5b178517361a Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Torma=20Krist=C3=B3f?= <tormakristof@tormakristof.eu>
Date: Sun, 3 Nov 2019 23:11:23 +0100
Subject: [PATCH] log processor script

---
 benchmark/classic/data/log_analyze.py | 86 +++++++++++++++++++++++++++
 benchmark/classic/data/process.py     |  4 +-
 benchmark/classic/wacher.sh           |  2 +-
 3 files changed, 89 insertions(+), 3 deletions(-)
 create mode 100644 benchmark/classic/data/log_analyze.py

diff --git a/benchmark/classic/data/log_analyze.py b/benchmark/classic/data/log_analyze.py
new file mode 100644
index 0000000..1c100ab
--- /dev/null
+++ b/benchmark/classic/data/log_analyze.py
@@ -0,0 +1,86 @@
+#!/usr/bin/env python3
+
+
+import json
+import datetime
+import re
+import pprint
+
+
+def average(lst: list) -> float:
+    return sum(lst) / len(lst)
+
+
+def listtodict(inlist: list) -> dict:
+    it = iter(inlist)
+    res_dct = dict(zip(it, it))
+    return res_dct
+
+
+def readfile() -> list:
+    lines = []
+    with open("log.txt", 'r') as inputFile:
+        line = inputFile.readline()
+        while line:
+            line = inputFile.readline()
+            try:
+                linedict = json.loads(line)
+                lines.append(linedict)
+            except json.JSONDecodeError:
+                continue
+    return lines
+
+
+def readconfigdates() -> dict:
+    dates = {}
+    with open("dates.txt", 'r') as inputFile:
+        line = inputFile.readline()
+        currline = 0
+        while line:
+            line = inputFile.readline()
+            currline += 1
+            dateformatted = datetime.datetime.strptime(
+                line, '%Y-%m-%d %H:%M:%S')
+            dates[currline] = dateformatted
+    return dates
+
+
+def procesfile(file: list, start: datetime, end: datetime) -> dict:
+    dictofsecs = {}
+    for line in file:
+        try:
+            currdate = line['ts'].split('.')[0].replace('T', ' ')
+            dateformatted = datetime.datetime.strptime(
+                currdate, '%Y-%m-%d %H:%M:%S')
+            if start < dateformatted < end:
+                message = line['msg']
+                messagelist = re.split(' |=', message)
+                messagedict = listtodict(messagelist)
+                messagedict['ts'] = dateformatted
+                if 'ObservedStableValue' in messagedict:
+                    if messagedict['ts'] not in dictofsecs:
+                        dictofsecs[messagedict['ts']] = []
+                        dictofsecs[messagedict['ts']].append(
+                            float(messagedict['ObservedStableValue']))
+                    else:
+                        dictofsecs[messagedict['ts']].append(
+                            float(messagedict['ObservedStableValue']))
+        except Exception as exception:
+            print(exception)
+    return dictofsecs
+
+
+def averagepersec(dictoftimes: dict, shouldprint: bool = False) -> dict:
+    dictofsecs = {}
+    for key, value in dictoftimes.items():
+        dictofsecs[key] = average(value)
+        if shouldprint:
+            print(dictofsecs[key])
+    return dictofsecs
+
+
+if __name__ == "__main__":
+    filelines = readfile()
+    config = readconfigdates()
+    fitdates = procesfile(filelines, config[0], config[1])
+    averageofdict = averagepersec(fitdates, True)
diff --git a/benchmark/classic/data/process.py b/benchmark/classic/data/process.py
index f5b283f..b1ae051 100644
--- a/benchmark/classic/data/process.py
+++ b/benchmark/classic/data/process.py
@@ -42,7 +42,7 @@ def processFile(fname):
             #print("Minimum response time was ", minResponse)
         else:
             print("csv is empty")
-        #pprint(responseCodes)
+        # pprint(responseCodes)
         for sec in responsePerSec:
             if len(responsePerSec[sec]) != 0:
                 #print(sec, ":")
@@ -50,7 +50,7 @@ def processFile(fname):
                 #print("    Minimum:", min(responsePerSec[sec]))
                 #print("    Num of responses:", len(responsePerSec[sec]))
                 print(len(responsePerSec[sec]))
-            #else:
+            # else:
                 #print("    empty")
 
 
diff --git a/benchmark/classic/wacher.sh b/benchmark/classic/wacher.sh
index 04af65d..371b695 100644
--- a/benchmark/classic/wacher.sh
+++ b/benchmark/classic/wacher.sh
@@ -5,6 +5,6 @@ function=hello-kubeless-go-sc
 
 for (( i = 1 ; i <= $time ; i++ )); do
   kubectl get deployment | grep $function | awk '{print $2}' | cut -d/ -f1 >> ./data/$function.monitor.csv
-  echo '\n' >> ./data/$function.monitor.csv
+#  echo '\n' >> ./data/$function.monitor.csv
   sleep 1
 done