Ping 99% Latencies

plot99.py

#!/usr/bin/env python

import argparse
import datetime
import matplotlib as mpl
mpl.use('Agg')
import matplotlib.pyplot as plt
import numpy
import os
import re
import sys
import time

def total_seconds(td):
  return td.days*24*60*60 + td.seconds

def parse(file):
  """
  Parse through the ping file and get the ping values.
  :param file: The input file for ping data.
  :return (x, y): The tuple containing a list of x values which are seconds since start and list y of ping values.
  """
  x = []
  y = []
  start = None
  for line in file:
    # Verify format has a time=#### ms
    m = re.search("(\d+-\d+-\d+ \d+:\d+:\d+).*time=([0-9\.]+) ms", line)
    if not m: continue

    ts = m.group(1)
    ts = datetime.datetime.strptime(ts, '%Y-%m-%d %H:%M:%S')
    y_i = float(m.group(2))

    if not start:
      start = ts
      x_i = 0
    else:
      x_i = total_seconds(ts - start)

    x.append(x_i)
    y.append(y_i)

  return (x, y)

def process_99(x, y):
  """
  Group the values in minute buckets and get 99% for each bucket.
  :param x: The seconds values.
  :param y: The actual ping values.
  :return (nx, ny): The tuple containing a list of nx values which are seconds since start and list ny of 99% ping values.
  """
  # Group into minute intervals and calculate 99%
  nx = []
  ny = []

  group_x = []
  group_y = []
  for (x_i, y_i) in zip(x, y):
    if len(group_x) == 0:
      group_x.append(x_i)
      group_y.append(y_i)
      continue

    if (x_i - group_x[0]) <= 60:
      # Within minute interval so group
      group_x.append(x_i)
      group_y.append(y_i)
    else:
      # Process the past minute and start new
      x_99 = group_x[0]  # Use first data point in minute section as data point
      y_99 = numpy.percentile(group_y, 99)

      nx.append(x_99)
      ny.append(y_99)
      # Start new group
      group_x = [x_i]
      group_y = [y_i]

  return (nx, ny)

if __name__ == "__main__":
  parser = argparse.ArgumentParser(description='Plot 99% ping latencies.')
  parser.add_argument('-i', '--input', type=str, required=True, help='the ping output file')

  args = parser.parse_args()
  input_filename = args.input

  if not os.path.isfile(input_filename):
    parser.print_help()
    sys.exit("Error: File does not exist, please provide a valid file.")

  # Process the file
  input_file = open(input_filename, 'r')
  (x, y) = parse(input_file)
  input_file.close()

  (nx, ny) = process_99(x, y)

  # Plot the data
  fig = plt.figure()
  ax = fig.add_subplot(111)
  ax.set_title("99 Percentile Ping Latencies")
  ax.set_xlabel("Seconds Since Collection (sec)")
  ax.set_ylabel("Ping (ms)")
  ax.plot(nx, ny)
  fig.savefig(input_filename + ".png")