#!/usr/bin/env python

# Copyright (c) 2011 Stanford University

#

# Permission to use, copy, modify, and distribute this software for any

# purpose with or without fee is hereby granted, provided that the above

# copyright notice and this permission notice appear in all copies.

#

# THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR(S) DISCLAIM ALL WARRANTIES

# WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF

# MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL AUTHORS BE LIABLE FOR

# ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES

# WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN

# ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF

# OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.

"""Generates data for a recovery performance graph."""

from __future__ import division, print_function

from common import *

import config

from glob import glob

import math

import metrics

import recovery

import re

import sys

def median(l):

l = sorted(l)

if len(l) % 2 == 0:

return metrics.average(l[len(l)//2:len(l)//2+1])

else:

return l[len(l)//2]

if len(sys.argv) > 1:

recovery_dir = sys.argv[1]

else:

recovery_dir = 'logs/latest'

NUMBACKUPS = len(config.hosts)

TRIALS = 25

backups = [[] for i in range(NUMBACKUPS)]

for trial in range(TRIALS):

args = {}

args['num_servers'] = NUMBACKUPS

args['backups_per_server'] = 1

args['num_partitions'] = 12

args['object_size'] = 1024

args['replicas'] = 3

args['num_objects'] = (626012 * args['num_servers'] * 80 //

args['num_partitions'] // 640)

r = recovery.insist(**args)

print('->', r['ns'] / 1e6, 'ms', '(run %s)' % r['run'])

for i, backup in enumerate(r['metrics'].backups):

backups[i].append(

(backup.backup.storageReadBytes / 2**20) /

(backup.backup.storageReadTicks / backup.clockFrequency))

with open('%s/recovery/disk_variability.data' % top_path, 'w', 1) as dat:

for outliersPass in [False, True]:

for i, read in enumerate(sorted(backups, key=median)):

read.sort()

m = median(read)

# Mendenhall and Sincich method

q1 = read[int(math.ceil(1 * (len(read) + 1) / 4)) - 1]

q3 = read[int(math.floor(3 * (len(read) + 1) / 4)) - 1]

iqr = q3 - q1

include = []

outliers = []

for x in read:

if x < q1 - 1.5 * iqr or x > q3 + 1.5 * iqr:

outliers.append(x)

else:

include.append(x)

if outliersPass:

for outlier in outliers:

print(i, outlier, file=dat)

else:

print(i, min(include), q1, m, q3, max(include), file=dat)

print(file=dat)

print(file=dat)