"伟大 (The Great)"的减速

由于许多国家和 ICP 供应商倾向于通过施加工具来控制或监视用户以达到优化他们“内部”数据访问的目的，致使整体网络状况恶化，性能明显下降。这个现状大大增加了为终端用户提供应用的公司的运营成本。我们的研究显示互联网的速度 — 不仅仅是在中国 — 在全球范围内都有着显著的差异并常常因为互联网没有选择优化路径而导致一些应用或一些用户短期或长期的访问缓慢。GrandeNet 可以通过使用 Re6st 在所有相连的服务器之间保持优质并稳定的连接路径，达到优化一些服务器80%的连接。

你会 Re6st 吗?

Re6st 是一个多协议随机网格生成器，使用 Babel 路由协议来发现网格中每个节点之间的最优路径。它可支持 IPv6 和 IPv4，也即将实现 RINA 支持。它被商业化应用于 VIFIB －我们的分布式云供应商，用来解决因路由器上的漏洞，通过数据包检测来阻拦TCP协议，严格的政府信息过滤系统等造成的企业级分布式应用面临的互联网连接缺乏可靠性的问题。没有 re6st, 是不可能在一个分布式云上部署大型企业（如三菱， Sanef，Aide et Action等）所需的关键商业应用。也不可能在巴西，中国或科特迪瓦这样缺乏可靠互联网环境的地方部署分布式云。

IPython Notebook

IPython Notebook 是一个基于网页的互动式计算环境, 用来创建嵌入 Python 代码的可执行 Notebooks。IPython Notebook 被研究人员大量用于研发并分享他们的科学成果。我们选择使用 IPython Notebook 在这篇文章中提供一个清晰的攻略。

这篇文章完全可以通过在你的 IPython Notebook 实例上导入这个 notebook 来重现。

为了执行这个 notebook，我们需要一些常见的 python 库，特别是 panda, numpy, scipy 和 matplotlib。以下是一些初始化库需要的导入。

In [1]:

%pylab inline
import sys
from matplotlib import pyplot
import matplotlib.pyplot as plt
#from mpl_toolkits.basemap import Basemap
from IPython.display import display, clear_output
from IPython.core.display import HTML 

import pandas as pd
import numpy as np

from pandas import rolling_median

np.set_printoptions(threshold="nan")

pd.set_option('display.max_rows', 2000)

from pandas import Series, DataFrame, Panel

Populating the interactive namespace from numpy and matplotlib

接下来，我们将定义核心代码－以方法的形式记录，将为这篇文章执行数据收集并计算结果。如果你对代码不感兴趣，你可以直接跳到这篇文章的下一章节。

In [2]:

import urllib2

def load_json(url, average="internet_ipv4", packet_lost="internet_ipv4_packet_lost"):
  """ Download JSON and normalize the data by:
        - Replace failed by a number '0', so the column is a float and not string
        - Replace 'average' and packet_lost by another name 
            for help concat 2 numpy array w/o recreate it.
        - Remove latest empty line.
  """
  req = urllib2.Request(url)
  response = urllib2.urlopen(req)
  content = response.read()
  return '[%s]' % content\
    .replace('"failed"', "0")\
    .replace('"average"', '"%s"' % average)\
    .replace('"packet_lost"', '"%s"' % packet_lost)\
    .replace("\n", ",")[:-1]

def load_test(id, date_id):
  """ Load Test results from Distributed Monitoring Tool
      and transform into DataFrames """
    
  # Load JSON for ICMPv6
  ping6_as_jsonstring = load_json(
    log_dict[id]["grandenet_ipv6"] % date_id, 
    average="grandenet_ipv6",  packet_lost="grandenet_ipv6_packet_lost")

  # Load JSON for ICMPv4  
  ping_as_jsonstring = load_json(
    log_dict[id]["internet_ipv4"] % date_id, 
    average="internet_ipv4", packet_lost="internet_ipv4_packet_lost")

  return pd.read_json(ping6_as_jsonstring, convert_dates=["time"]), \
    pd.read_json(ping_as_jsonstring, convert_dates=["time"])

def get_computer_list(dframeA, dframeB ):
    """ Extract all computer names at the DataFrames """
    return list(set([ computer_name[0] for computer_name in dframeA[["computer_name"]].as_matrix()] + 
                  [ computer_name[0] for computer_name in dframeB[["computer_name"]].as_matrix()]))

def get_computer_destination_label(dframeA):
    """ Determinate the Label Name for the computer which are receiving the ping"""
    return getComputerLabel([computer_name[0] for computer_name in dframeA[["name_or_ip"]].as_matrix()][0])

def getComputerLabel(computer_name):
    """ Translate hostname, ip addresses into meaningfull names for better understanting"""
    return server_label.get(computer_name, computer_name)

# Initiallization function which are going to be used for 
# collect the logs and transform them on Data Frames.
def plot_ping_comparation(df_ping6, df_ping):
  """ Function to load, plot and compare 2 Data Frames 
  """
  computer_list = get_computer_list(df_ping, df_ping6)

  computer_destination_label = get_computer_destination_label(df_ping6)

  measured_average = []
  packet_lost = []
    
  for computer_name in computer_list:
    
    if getComputerLabel(computer_name) == computer_destination_label:
      continue

    df6 = pd.DataFrame(df_ping6[df_ping6["computer_name"] == computer_name][df_ping6["grandenet_ipv6"] > 0][["time", "grandenet_ipv6"]])    
    df4 = pd.DataFrame(df_ping[df_ping["computer_name"] == computer_name][df_ping["internet_ipv4"] > 0][["time", "internet_ipv4"]])

    # Use Moving average in order to eliminate noise spikes on the chart and measurement.
    df6['grandenet_ipv6'] = rolling_median(df6['grandenet_ipv6'], window=3, center=True)
    df4['internet_ipv4'] = rolling_median(df4['internet_ipv4'], window=3, center=True)
    
    label = "'%s' to '%s'" % (getComputerLabel(computer_name), computer_destination_label)

    if 0 in [len(df6), len(df4)]:
      print "Found one empty array for %s" % label
      continue
    
    df = pd.DataFrame(pd.concat([df6, df4]))

    if SHOW_ALL_CHARTS:
      df4.plot(x="time", title=label + " (lower is better)", 
               sort_columns=["time"], figsize=(20,6))
      df6.plot(x="time", title=label + " (lower is better)", 
               sort_columns=["time"], color='r', figsize=(20,6))
    
    df.plot(x="time", title=label + " (lower is better)",
            marker='o', color=["b", "r"], figsize=(20,6))
    
    # Ignore 0 entries as it represents a full failure (so no average).
    ipv6_mean = df6["grandenet_ipv6"].mean()
    ipv4_mean = df4["internet_ipv4"].mean()

    grandenet_ipv6_packet_lost = df_ping6[df_ping6["computer_name"] == computer_name]["grandenet_ipv6_packet_lost"].mean()
    internet_ipv4_packet_lost = df_ping[df_ping["computer_name"] == computer_name]["internet_ipv4_packet_lost"].mean()
 
    if ipv6_mean < ipv4_mean:
      improvement_ratio = float(ipv4_mean - ipv6_mean)/ipv4_mean
      state = "OPTIMIZED in %sms (%.2f%%)" % ((ipv4_mean - ipv6_mean), improvement_ratio*100)
    elif ipv6_mean < (ipv4_mean + max(20, ipv4_mean*0.15)):
      state = "OK (in acceptable range %s < %s < %s)" % (ipv4_mean, ipv6_mean, (ipv4_mean + max(20, ipv4_mean*0.15)))
    else:
      state = "BAD (%sms slower)" % (ipv6_mean - ipv4_mean)

    measured_average.append({"name" : "'%s' to '%s'" % (getComputerLabel(computer_name), computer_destination_label),
                             "grandenet_ipv6": ipv6_mean,
                             "internet_ipv4": ipv4_mean,
                             "state": state})

    
    if grandenet_ipv6_packet_lost < internet_ipv4_packet_lost:
      loss_state = "OPTIMIZED (Better Packet Lost rate)"
    elif grandenet_ipv6_packet_lost == internet_ipv4_packet_lost:
      loss_state = "OK (Same Packet Lost rate)"
    elif (grandenet_ipv6_packet_lost - internet_ipv4_packet_lost) < 1:
      loss_state = "OK (less them 1% diference is considered same)"
    else:
      loss_state = "BAD (Worst Packet Lost rate)"

    packet_lost.append({"name" : "'%s' to '%s'" % (getComputerLabel(computer_name), computer_destination_label),
                             "grandenet_ipv6_packet_lost": grandenet_ipv6_packet_lost,
                             "internet_ipv4_packet_lost": internet_ipv4_packet_lost,
                             "state": loss_state})

  return pd.DataFrame(measured_average), pd.DataFrame(packet_lost)

用 SlapOS 分布式监测评估性能

GrandeNet 架构的核心是基于分布在多个云供应商 (亚马逊，青云，OVH，Rackspace，Ucloud 等) 服务器上以及分布在企业办公区或个人家庭的独立机器上。客户可以添加位于他们处所甚至是家庭的服务器，来用作主要的生产服务器。

这个 GrandeNet 混合动力和异构基础架构使用 SlapOS 管理和监测全球所有的分布式服务器。

在这篇文章中，我们使用一小组服务器 (12个) 使用 IPv4 运行 SlapOS 分布式监测。每个服务器都试图联系 (使用 ICMP 协议）使用 IPv4 和 IPv6 的12个服务器的地址。每10分钟运行10次测试 (10 pings) ，然后我们来测试并比较观察到的平均值和数据流失量。

下面这个图像说明了仅使用3个服务器时的测试:

GrandeNet Connectivity Example

下面我们初始化了每个服务器日志的位置，以及标签来提高图标和结果的可读性。

In [3]:

server_label = {
  'i-j0dshts2': "Guanghouz - Qincloud",
  '10-13-16-6': "Guanghouz - UCloud",
  'i-wbs0d67i' : "Hongkong - Qincloud 1",
  'i-vutfghrs': "Hongkong - Qincloud 0",
  'i-hf0f7ocn': "Beijing - Qincloud",
  'vps212661.ovh.net': "Strasbourg - OVH",
  'ip-172-31-30-97': "Singapour - Amazon",
  'ip-172-31-6-206': "Tokyo - Amazon",
  'ip-172-31-8-66' : "Virginia - Amazon",
  'ip-172-31-7-155': "US West - Amazon",
  'cloud-server-grandenet' : "Hongkong - Rackspace",
    
  'COMP-9': 'US West - Amazon',
  'COMP-8': 'Singapour - Amazon',
  'COMP-7': 'Tokyo - Amazon',
  'COMP-6': 'Hongkong - Qincloud 1',
  'COMP-4': 'Hongkong - Qincloud 0',
  'COMP-2': 'Beijing - Qincloud',
  'COMP-3': 'Guanghouz - Qincloud',
  'COMP-10': 'Guanghouz - UCloud',
  'COMP-11': 'Strasbourg - OVH',
  'COMP-12': 'Virginia - Amazon',
  "COMP-13": 'Beauharnois - OVH',

  "frontend0.grandenet.cn": "Beijing - Qincloud",
  "2401:5180::1": "Beijing - Qincloud",
  "frontend1.grandenet.cn": "Guanghouz - Qincloud",
  "2401:5180:0:6::1": "Guanghouz - Qincloud",
  "2401:5180:0:9::1" : "Hongkong - Qincloud 0",
  "frontend3.grandenet.cn" : "Hongkong - Qincloud 0",
  "2401:5180:0:8::1" : "Hongkong - Rackspace",
  "frontend4.grandenet.cn" : "Hongkong - Rackspace",
  "2401:5180:0:7::1": "Hongkong - Qincloud 1",
  "frontend5.grandenet.cn": "Hongkong - Qincloud 1",
  "2401:5180:0:c::1": "Tokyo - Amazon", 
  "frontend7.grandenet.cn": "Tokyo - Amazon",
  "2401:5180:0:d::1": "Singapour - Amazon",
  "frontend6.grandenet.cn": "Singapour - Amazon",
  "2401:5180:0:10::1": "US West - Amazon",
  "frontend8.grandenet.cn": "US West - Amazon",
  "2401:5180:0:13::1": "Guanghouz - UCloud",
  "frontend9.grandenet.cn": "Guanghouz - UCloud",
  "2401:5180:0:16::1": "Strasbourg - OVH",
  "frontend10.grandenet.cn": "Strasbourg - OVH",
  "2401:5180:0:15::1": "Virginia - Amazon",
  "frontend11.grandenet.cn": "Virginia - Amazon",
  "2401:5180:0:17::1": "Beauharnois - OVH",
  "frontend12.grandenet.cn": "Beauharnois - OVH",

}


log_dict = {
  "Hongkong - Qincloud 0": {
      "grandenet_ipv6": "https://softinst303.node.grandenet.cn/SOFTINST-314/ping6/log.%s.log",
      "internet_ipv4": "https://softinst303.node.grandenet.cn/SOFTINST-314/ping/log.%s.log",
      },
  "Virginia - Amazon": {
      "grandenet_ipv6": "https://softinst303.node.grandenet.cn/SOFTINST-322/ping6/log.%s.log",
      "internet_ipv4": "https://softinst303.node.grandenet.cn/SOFTINST-322/ping/log.%s.log",
      },
  "Strasbourg - OVH": {
      "grandenet_ipv6": "https://softinst303.node.grandenet.cn/SOFTINST-321/ping6/log.%s.log",
      "internet_ipv4": "https://softinst303.node.grandenet.cn/SOFTINST-321/ping/log.%s.log",
      },
  "Guanghouz - UCloud": {
      "grandenet_ipv6": "https://softinst303.node.grandenet.cn/SOFTINST-320/ping6/log.%s.log",
      "internet_ipv4": "https://softinst303.node.grandenet.cn/SOFTINST-320/ping/log.%s.log",
      },
  "Tokyo - Amazon": {
      "grandenet_ipv6": "https://softinst303.node.grandenet.cn/SOFTINST-317/ping6/log.%s.log",
      "internet_ipv4": "https://softinst303.node.grandenet.cn/SOFTINST-317/ping/log.%s.log",
      },
  "US West - Amazon": {
      "grandenet_ipv6": "https://softinst303.node.grandenet.cn/SOFTINST-319/ping6/log.%s.log",
      "internet_ipv4": "https://softinst303.node.grandenet.cn/SOFTINST-319/ping/log.%s.log",
      },
  "Singapour - Amazon": {
      "grandenet_ipv6": "https://softinst303.node.grandenet.cn/SOFTINST-318/ping6/log.%s.log",
      "internet_ipv4": "https://softinst303.node.grandenet.cn/SOFTINST-318/ping/log.%s.log",
      },
  "Hongkong - Qincloud 1": {
      "grandenet_ipv6": "https://softinst303.node.grandenet.cn/SOFTINST-316/ping6/log.%s.log",
      "internet_ipv4": "https://softinst303.node.grandenet.cn/SOFTINST-316/ping/log.%s.log",
      },
  "Guanghouz - Qincloud": {
      "grandenet_ipv6": "https://softinst303.node.grandenet.cn/SOFTINST-313/ping6/log.%s.log",
      "internet_ipv4": "https://softinst303.node.grandenet.cn/SOFTINST-313/ping/log.%s.log",
      },
  "Beijing - Qincloud": {
      "grandenet_ipv6": "https://softinst303.node.grandenet.cn/SOFTINST-312/ping6/log.%s.log",
      "internet_ipv4": "https://softinst303.node.grandenet.cn/SOFTINST-312/ping/log.%s.log",
      },
  "Hongkong - Rackspace": {
      "grandenet_ipv6": "https://softinst303.node.grandenet.cn/SOFTINST-315/ping6/log.%s.log",
      "internet_ipv4": "https://softinst303.node.grandenet.cn/SOFTINST-315/ping/log.%s.log",
      },
  "Beauharnois - OVH": {
      "grandenet_ipv6": "https://softinst303.node.grandenet.cn/SOFTINST-625/ping6/log.%s.log",
      "internet_ipv4": "https://softinst303.node.grandenet.cn/SOFTINST-625/ping/log.%s.log",
      }
}

我们同时还限制了这篇文章的规模，使测试运行在一定日期范围内，见如下变量“DAY”。

In [4]:

# Define here if you want more or less charts verbosity. Show all charts can
# make this report quite big.
SHOW_ALL_CHARTS = False

# Generate Report for the Jan, 28, 2016
DAY = "20160128"

从分布式 SlapOS 监测收集数据

为了出结果，我们使用了如上定义的方法，抓取日志并转化为数据帧。这些数据帧包含了指示时期 (DAY) 的测试结果。

In [5]:

hq0_df_ping6, hq0_df_ping = load_test(id = "Hongkong - Qincloud 0", date_id=DAY)

In [6]:

va_df_ping6, va_df_ping = load_test(id = "Virginia - Amazon", date_id=DAY)

In [7]:

gu_df_ping6, gu_df_ping = load_test(id =  "Guanghouz - UCloud", date_id=DAY)

In [8]:

sa_df_ping6, sa_df_ping = load_test(id = "Singapour - Amazon", date_id=DAY)

In [9]:

hq1_df_ping6, hq1_df_ping = load_test(id = "Hongkong - Qincloud 1", date_id=DAY)

In [10]:

hr_df_ping6, hr_df_ping = load_test(id = "Hongkong - Rackspace", date_id=DAY)

In [11]:

wa_df_ping6, wa_df_ping = load_test(id = "US West - Amazon", date_id=DAY)

In [12]:

go_df_ping6, go_df_ping = load_test(id = "Strasbourg - OVH", date_id=DAY)

In [13]:

ta_df_ping6, ta_df_ping = load_test(id = "Tokyo - Amazon", date_id=DAY)

In [14]:

gq_df_ping6, gq_df_ping = load_test(id = "Guanghouz - Qincloud", date_id=DAY)

In [15]:

bq_df_ping6, bq_df_ping = load_test(id = "Beijing - Qincloud", date_id=DAY)

In [16]:

bho_df_ping6, bho_df_ping = load_test(id = "Beauharnois - OVH", date_id=DAY)

互联网 IPv4 vs Grandenet IPv6

使用数据帧，我们可以视觉化在使用互联网 IPv4 (红色）和 Grandenet IPv6 (蓝色) 时的响应时间（毫秒）的比较。由于我们使用的是 ICMP 协议来测量响应时间，下面的图表使用了“ping"作为 IPv4的名字，ping6 作为 IPv6 的名字，并且突出了互联网 IPv4 和 IPv6 之间的区别。响应时间越短，绘制线越低，越好。

In [17]:

hq0_average_dataframe, hq0_packetloss_dataframe = plot_ping_comparation(hq0_df_ping6, hq0_df_ping)

/srv/slapgrid/slappart8/srv/runner/software/9a8d67b31671ba36ac107c65a141c073/develop-eggs/pandas-0.16.2-py2.7-linux-x86_64.egg/pandas/core/frame.py:1825: UserWarning: Boolean Series key will be reindexed to match DataFrame index.
  "DataFrame index.", UserWarning)

In [18]:

va_average_dataframe, va_packetloss_dataframe = plot_ping_comparation(va_df_ping6, va_df_ping)

In [19]:

go_average_dataframe, go_packetloss_dataframe = plot_ping_comparation(go_df_ping6, go_df_ping)

In [20]:

gu_average_dataframe, gu_packetloss_dataframe = plot_ping_comparation(gu_df_ping6, gu_df_ping)

In [21]:

ta_average_dataframe, ta_packetloss_dataframe = plot_ping_comparation(ta_df_ping6, ta_df_ping)

In [22]:

sa_average_dataframe, sa_packetloss_dataframe = plot_ping_comparation(sa_df_ping6, sa_df_ping)

In [23]:

hq1_average_dataframe, hq1_packetloss_dataframe = plot_ping_comparation(hq1_df_ping6, hq1_df_ping)

In [24]:

gq_average_dataframe, gq_packetloss_dataframe = plot_ping_comparation(gq_df_ping6, gq_df_ping)

In [25]:

bq_average_dataframe, bq_packetloss_dataframe = plot_ping_comparation(bq_df_ping6, bq_df_ping)

In [26]:

wa_average_dataframe, wa_packetloss_dataframe = plot_ping_comparation(wa_df_ping6, wa_df_ping)

In [27]:

hr_average_dataframe, hr_packetloss_dataframe = plot_ping_comparation(hr_df_ping6, hr_df_ping)

In [28]:

bho_average_dataframe, bho_packetloss_dataframe = plot_ping_comparation(bho_df_ping6, bho_df_ping)

IPv4 和 IPv6 地址的平均 Ping 响应比较

我们都知道网络延迟对网页应用的性能有直接的影响。通过比较 IPv4 和 IPv6 的平均 ping 响应时间，你会注意到服务器性能的极大提升。

通过使用 Grandenet, 我们来看两个服务器之间连接的3个状态：

当 IPv6 的速度比 IPv4 的速度快时，连接质量就被优化了
当 IPv4 比 IPv6 快超过15%或 20毫秒时，连接质量就不太好
当 IPv4 比 IPv6 快不超过15%或20毫秒时，连接质量就还不错

这些状态承认了一个事实，即在最坏的情况下，re6st 也可以提升 ping 响应时间至少20秒，也就是两个相隔很远的两个服务器之间直接相连。

一个理想的情景是在做相互测试的时候让所有的服务器都处于被优化或还不错的状态。下面的表格就包含了之前所提到的服务器之间的全部连接列表。你可以发现不仅仅是有中国服务器的连接被优化了，日本和新加坡，日本和香港，美国和香港以及其他的连接都被优化了。

In [29]:

df_average_statistic = pd.concat([go_average_dataframe, hr_average_dataframe, 
                                  bho_average_dataframe,bq_average_dataframe, 
                                  gq_average_dataframe, va_average_dataframe, 
                                  hq0_average_dataframe, gu_average_dataframe,  
                                  ta_average_dataframe, sa_average_dataframe, 
                                  hq1_average_dataframe, wa_average_dataframe],
                                 ignore_index=True)

df_average_statistic

Out[29]:

	grandenet_ipv6	internet_ipv4	name	state
0	230.769642	227.776942	'Beijing - Qincloud' to 'Strasbourg - OVH'	OK (in acceptable range 227.776942149 < 230.76...
1	207.231196	199.779662	'Singapour - Amazon' to 'Strasbourg - OVH'	OK (in acceptable range 199.779661871 < 207.23...
2	234.477413	343.148734	'Hongkong - Qincloud 1' to 'Strasbourg - OVH'	OPTIMIZED in 108.671320769ms (31.67%)
3	235.069314	344.405986	'Hongkong - Qincloud 0' to 'Strasbourg - OVH'	OPTIMIZED in 109.336671743ms (31.75%)
4	159.144745	160.710180	'US West - Amazon' to 'Strasbourg - OVH'	OPTIMIZED in 1.56543533057ms (0.97%)
5	121.403561	121.298650	'Virginia - Amazon' to 'Strasbourg - OVH'	OK (in acceptable range 121.29865 < 121.403561...
6	246.411341	356.375993	'Guanghouz - Qincloud' to 'Strasbourg - OVH'	OPTIMIZED in 109.964652226ms (30.86%)
7	90.044151	90.002443	'Beauharnois - OVH' to 'Strasbourg - OVH'	OK (in acceptable range 90.0024428571 < 90.044...
8	254.088217	372.031417	'Guanghouz - UCloud' to 'Strasbourg - OVH'	OPTIMIZED in 117.943199875ms (31.70%)
9	247.874190	325.955576	'Hongkong - Rackspace' to 'Strasbourg - OVH'	OPTIMIZED in 78.0813857585ms (23.95%)
10	266.606167	278.157007	'Tokyo - Amazon' to 'Strasbourg - OVH'	OPTIMIZED in 11.5508405276ms (4.15%)
11	243.557295	325.525186	'Strasbourg - OVH' to 'Hongkong - Rackspace'	OPTIMIZED in 81.9678907503ms (25.18%)
12	41.762925	39.811375	'Beijing - Qincloud' to 'Hongkong - Rackspace'	OK (in acceptable range 39.811375 < 41.762925 ...
13	38.099014	34.921950	'Singapour - Amazon' to 'Hongkong - Rackspace'	OK (in acceptable range 34.9219496403 < 38.099...
14	2.153391	1.995842	'Hongkong - Qincloud 1' to 'Hongkong - Rackspace'	OK (in acceptable range 1.99584172662 < 2.1533...
15	2.160094	1.965309	'Hongkong - Qincloud 0' to 'Hongkong - Rackspace'	OK (in acceptable range 1.96530935252 < 2.1600...
16	164.382985	157.351396	'US West - Amazon' to 'Hongkong - Rackspace'	OK (in acceptable range 157.351395683 < 164.38...
17	211.392151	229.954993	'Virginia - Amazon' to 'Hongkong - Rackspace'	OPTIMIZED in 18.562841778ms (8.07%)
18	8.568826	8.248223	'Guanghouz - Qincloud' to 'Hongkong - Rackspace'	OK (in acceptable range 8.24822302158 < 8.5688...
19	229.472014	428.977350	'Beauharnois - OVH' to 'Hongkong - Rackspace'	OPTIMIZED in 199.505335612ms (46.51%)
20	12.660536	7.320137	'Guanghouz - UCloud' to 'Hongkong - Rackspace'	OK (in acceptable range 7.32013669065 < 12.660...
21	56.437174	71.705446	'Tokyo - Amazon' to 'Hongkong - Rackspace'	OPTIMIZED in 15.2682721301ms (21.29%)
22	90.059710	89.985175	'Strasbourg - OVH' to 'Beauharnois - OVH'	OK (in acceptable range 89.9851751825 < 90.059...
23	254.805042	288.993650	'Beijing - Qincloud' to 'Beauharnois - OVH'	OPTIMIZED in 34.1886079832ms (11.83%)
24	247.255138	269.330863	'Singapour - Amazon' to 'Beauharnois - OVH'	OPTIMIZED in 22.0757256282ms (8.20%)
25	224.304181	251.024281	'Hongkong - Qincloud 1' to 'Beauharnois - OVH'	OPTIMIZED in 26.7200994161ms (10.64%)
26	226.983348	284.750353	'Hongkong - Qincloud 0' to 'Beauharnois - OVH'	OPTIMIZED in 57.7670046919ms (20.29%)
27	70.884467	70.894246	'US West - Amazon' to 'Beauharnois - OVH'	OPTIMIZED in 0.00977922352695ms (0.01%)
28	43.673137	42.949250	'Virginia - Amazon' to 'Beauharnois - OVH'	OK (in acceptable range 42.94925 < 43.67313669...
29	232.296942	329.106576	'Guanghouz - Qincloud' to 'Beauharnois - OVH'	OPTIMIZED in 96.8096335106ms (29.42%)
30	237.192766	385.034875	'Guanghouz - UCloud' to 'Beauharnois - OVH'	OPTIMIZED in 147.842108577ms (38.40%)
31	228.411906	423.219410	'Hongkong - Rackspace' to 'Beauharnois - OVH'	OPTIMIZED in 194.807504275ms (46.03%)
32	177.716043	184.348388	'Tokyo - Amazon' to 'Beauharnois - OVH'	OPTIMIZED in 6.63234501095ms (3.60%)
33	240.231168	235.254507	'Strasbourg - OVH' to 'Beijing - Qincloud'	OK (in acceptable range 235.254507143 < 240.23...
34	76.299868	70.598232	'Singapour - Amazon' to 'Beijing - Qincloud'	OK (in acceptable range 70.5982318841 < 76.299...
35	39.032730	41.159784	'Hongkong - Qincloud 1' to 'Beijing - Qincloud'	OPTIMIZED in 2.12705424565ms (5.17%)
36	44.635572	43.729914	'Hongkong - Qincloud 0' to 'Beijing - Qincloud'	OK (in acceptable range 43.7299136691 < 44.635...
37	191.950588	172.012237	'US West - Amazon' to 'Beijing - Qincloud'	OK (in acceptable range 172.012237037 < 191.95...
38	239.912219	236.548706	'Virginia - Amazon' to 'Beijing - Qincloud'	OK (in acceptable range 236.548705882 < 239.91...
39	46.273000	47.747281	'Guanghouz - Qincloud' to 'Beijing - Qincloud'	OPTIMIZED in 1.47428057554ms (3.09%)
40	261.261261	301.428700	'Beauharnois - OVH' to 'Beijing - Qincloud'	OPTIMIZED in 40.1674391304ms (13.33%)
41	43.210993	52.811683	'Guanghouz - UCloud' to 'Beijing - Qincloud'	OPTIMIZED in 9.60069069961ms (18.18%)
42	42.063942	41.977022	'Hongkong - Rackspace' to 'Beijing - Qincloud'	OK (in acceptable range 41.9770215827 < 42.063...
43	96.826109	135.557187	'Tokyo - Amazon' to 'Beijing - Qincloud'	OPTIMIZED in 38.7310783547ms (28.57%)
44	253.286304	348.142114	'Strasbourg - OVH' to 'Guanghouz - Qincloud'	OPTIMIZED in 94.8558099379ms (27.25%)
45	45.542403	47.665650	'Beijing - Qincloud' to 'Guanghouz - Qincloud'	OPTIMIZED in 2.12324663866ms (4.45%)
46	46.796123	193.877259	'Singapour - Amazon' to 'Guanghouz - Qincloud'	OPTIMIZED in 147.081135804ms (75.86%)
47	11.600123	28.864662	'Hongkong - Qincloud 1' to 'Guanghouz - Qincloud'	OPTIMIZED in 17.2645386821ms (59.81%)
48	14.198486	37.674871	'Hongkong - Qincloud 0' to 'Guanghouz - Qincloud'	OPTIMIZED in 23.4763849964ms (62.31%)
49	173.533544	292.845978	'US West - Amazon' to 'Guanghouz - Qincloud'	OPTIMIZED in 119.3124343ms (40.74%)
50	220.941986	294.773657	'Virginia - Amazon' to 'Guanghouz - Qincloud'	OPTIMIZED in 73.8316716356ms (25.05%)
51	236.891568	361.344309	'Beauharnois - OVH' to 'Guanghouz - Qincloud'	OPTIMIZED in 124.452741007ms (34.44%)
52	5.721768	5.410108	'Guanghouz - UCloud' to 'Guanghouz - Qincloud'	OK (in acceptable range 5.41010791367 < 5.7217...
53	8.664124	8.320640	'Hongkong - Rackspace' to 'Guanghouz - Qincloud'	OK (in acceptable range 8.32064028777 < 8.6641...
54	70.081956	131.256727	'Tokyo - Amazon' to 'Guanghouz - Qincloud'	OPTIMIZED in 61.1747704143ms (46.61%)
55	121.487640	121.247236	'Strasbourg - OVH' to 'Virginia - Amazon'	OK (in acceptable range 121.247235714 < 121.48...
56	240.120950	230.511413	'Beijing - Qincloud' to 'Virginia - Amazon'	OK (in acceptable range 230.511413223 < 240.12...
57	222.681261	223.468777	'Singapour - Amazon' to 'Virginia - Amazon'	OPTIMIZED in 0.787516108852ms (0.35%)
58	209.687812	210.714964	'Hongkong - Qincloud 1' to 'Virginia - Amazon'	OPTIMIZED in 1.02715243457ms (0.49%)
59	206.991920	218.024158	'Hongkong - Qincloud 0' to 'Virginia - Amazon'	OPTIMIZED in 11.0322379835ms (5.06%)
60	76.374496	84.404000	'US West - Amazon' to 'Virginia - Amazon'	OPTIMIZED in 8.02950364964ms (9.51%)
61	220.761130	238.426374	'Guanghouz - Qincloud' to 'Virginia - Amazon'	OPTIMIZED in 17.6652436659ms (7.41%)
62	43.641381	42.894914	'Beauharnois - OVH' to 'Virginia - Amazon'	OK (in acceptable range 42.8949142857 < 43.641...
63	225.314246	389.168935	'Guanghouz - UCloud' to 'Virginia - Amazon'	OPTIMIZED in 163.854688875ms (42.10%)
64	211.599536	229.980266	'Hongkong - Rackspace' to 'Virginia - Amazon'	OPTIMIZED in 18.3807299552ms (7.99%)
65	152.523601	152.000986	'Tokyo - Amazon' to 'Virginia - Amazon'	OK (in acceptable range 152.000985612 < 152.52...
66	239.363849	347.087293	'Strasbourg - OVH' to 'Hongkong - Qincloud 0'	OPTIMIZED in 107.723443936ms (31.04%)
67	45.002513	41.264050	'Beijing - Qincloud' to 'Hongkong - Qincloud 0'	OK (in acceptable range 41.26405 < 45.00251260...
68	36.321246	35.532187	'Singapour - Amazon' to 'Hongkong - Qincloud 0'	OK (in acceptable range 35.5321870504 < 36.321...
69	0.837732	0.606295	'Hongkong - Qincloud 1' to 'Hongkong - Qinclou...	OK (in acceptable range 0.606294964029 < 0.837...
70	160.017212	169.572899	'US West - Amazon' to 'Hongkong - Qincloud 0'	OPTIMIZED in 9.55568760174ms (5.64%)
71	207.156094	218.038657	'Virginia - Amazon' to 'Hongkong - Qincloud 0'	OPTIMIZED in 10.8825636177ms (4.99%)
72	14.436101	36.616367	'Guanghouz - Qincloud' to 'Hongkong - Qincloud 0'	OPTIMIZED in 22.1802654572ms (60.57%)
73	229.083065	283.335871	'Beauharnois - OVH' to 'Hongkong - Qincloud 0'	OPTIMIZED in 54.2528066804ms (19.15%)
74	19.104543	63.614424	'Guanghouz - UCloud' to 'Hongkong - Qincloud 0'	OPTIMIZED in 44.5098809822ms (69.97%)
75	2.138971	1.951871	'Hongkong - Rackspace' to 'Hongkong - Qincloud 0'	OK (in acceptable range 1.9518705036 < 2.13897...
76	53.963688	51.621165	'Tokyo - Amazon' to 'Hongkong - Qincloud 0'	OK (in acceptable range 51.6211654676 < 53.963...
77	255.833453	365.678857	'Strasbourg - OVH' to 'Guanghouz - UCloud'	OPTIMIZED in 109.845403905ms (30.04%)
78	43.109625	52.778208	'Beijing - Qincloud' to 'Guanghouz - UCloud'	OPTIMIZED in 9.66858333333ms (18.32%)
79	52.838234	227.743525	'Singapour - Amazon' to 'Guanghouz - UCloud'	OPTIMIZED in 174.905291603ms (76.80%)
80	15.742594	69.543173	'Hongkong - Qincloud 1' to 'Guanghouz - UCloud'	OPTIMIZED in 53.800578459ms (77.36%)
81	20.123913	66.334302	'Hongkong - Qincloud 0' to 'Guanghouz - UCloud'	OPTIMIZED in 46.2103891148ms (69.66%)
82	178.273566	342.378604	'US West - Amazon' to 'Guanghouz - UCloud'	OPTIMIZED in 164.10503814ms (47.93%)
83	225.984885	382.897636	'Virginia - Amazon' to 'Guanghouz - UCloud'	OPTIMIZED in 156.912750822ms (40.98%)
84	5.724739	5.426122	'Guanghouz - Qincloud' to 'Guanghouz - UCloud'	OK (in acceptable range 5.42612230216 < 5.7247...
85	242.619297	421.211616	'Beauharnois - OVH' to 'Guanghouz - UCloud'	OPTIMIZED in 178.592318841ms (42.40%)
86	14.407255	7.411892	'Hongkong - Rackspace' to 'Guanghouz - UCloud'	OK (in acceptable range 7.41189208633 < 14.407...
87	71.360387	121.463964	'Tokyo - Amazon' to 'Guanghouz - UCloud'	OPTIMIZED in 50.1035771675ms (41.25%)
88	266.439424	278.120979	'Strasbourg - OVH' to 'Tokyo - Amazon'	OPTIMIZED in 11.681554111ms (4.20%)
89	95.726546	130.562108	'Beijing - Qincloud' to 'Tokyo - Amazon'	OPTIMIZED in 34.8355621148ms (26.68%)
90	74.860529	74.136101	'Singapour - Amazon' to 'Tokyo - Amazon'	OK (in acceptable range 74.1361007194 < 74.860...
91	57.655964	57.372640	'Hongkong - Qincloud 1' to 'Tokyo - Amazon'	OK (in acceptable range 57.3726402878 < 57.655...
92	54.828377	51.194863	'Hongkong - Qincloud 0' to 'Tokyo - Amazon'	OK (in acceptable range 51.1948633094 < 54.828...
93	108.687358	105.628540	'US West - Amazon' to 'Tokyo - Amazon'	OK (in acceptable range 105.628539568 < 108.68...
94	152.578827	152.403857	'Virginia - Amazon' to 'Tokyo - Amazon'	OK (in acceptable range 152.403857143 < 152.57...
95	70.801558	147.181209	'Guanghouz - Qincloud' to 'Tokyo - Amazon'	OPTIMIZED in 76.3796506621ms (51.89%)
96	178.120978	184.354986	'Beauharnois - OVH' to 'Tokyo - Amazon'	OPTIMIZED in 6.23400729702ms (3.38%)
97	74.521449	139.606777	'Guanghouz - UCloud' to 'Tokyo - Amazon'	OPTIMIZED in 65.0853277031ms (46.62%)
98	55.151833	73.062180	'Hongkong - Rackspace' to 'Tokyo - Amazon'	OPTIMIZED in 17.9103465228ms (24.51%)
99	205.582540	199.781236	'Strasbourg - OVH' to 'Singapour - Amazon'	OK (in acceptable range 199.781235714 < 205.58...
100	76.846342	70.112595	'Beijing - Qincloud' to 'Singapour - Amazon'	OK (in acceptable range 70.1125950413 < 76.846...
101	36.377196	35.139094	'Hongkong - Qincloud 1' to 'Singapour - Amazon'	OK (in acceptable range 35.1390935252 < 36.377...
102	36.252500	35.560604	'Hongkong - Qincloud 0' to 'Singapour - Amazon'	OK (in acceptable range 35.5606043165 < 36.252...
103	176.300474	175.301252	'US West - Amazon' to 'Singapour - Amazon'	OK (in acceptable range 175.301251799 < 176.30...
104	222.421165	223.566721	'Virginia - Amazon' to 'Singapour - Amazon'	OPTIMIZED in 1.14555596095ms (0.51%)
105	45.856601	184.241842	'Guanghouz - Qincloud' to 'Singapour - Amazon'	OPTIMIZED in 138.385240277ms (75.11%)
106	247.191863	269.351136	'Beauharnois - OVH' to 'Singapour - Amazon'	OPTIMIZED in 22.1592724049ms (8.23%)
107	50.988957	215.262813	'Guanghouz - UCloud' to 'Singapour - Amazon'	OPTIMIZED in 164.273856428ms (76.31%)
108	38.258377	34.939712	'Hongkong - Rackspace' to 'Singapour - Amazon'	OK (in acceptable range 34.9397122302 < 38.258...
109	74.865471	74.137086	'Tokyo - Amazon' to 'Singapour - Amazon'	OK (in acceptable range 74.1370863309 < 74.865...
110	238.208804	345.964200	'Strasbourg - OVH' to 'Hongkong - Qincloud 1'	OPTIMIZED in 107.755395652ms (31.15%)
111	38.542218	38.900467	'Beijing - Qincloud' to 'Hongkong - Qincloud 1'	OPTIMIZED in 0.358248179272ms (0.92%)
112	36.307116	35.202827	'Singapour - Amazon' to 'Hongkong - Qincloud 1'	OK (in acceptable range 35.2028273381 < 36.307...
113	1.923449	0.613331	'Hongkong - Qincloud 0' to 'Hongkong - Qinclou...	OK (in acceptable range 0.613330935252 < 1.923...
114	158.409131	163.004000	'US West - Amazon' to 'Hongkong - Qincloud 1'	OPTIMIZED in 4.59486861314ms (2.82%)
115	209.810568	210.780729	'Virginia - Amazon' to 'Hongkong - Qincloud 1'	OPTIMIZED in 0.970160226105ms (0.46%)
116	13.477819	30.705424	'Guanghouz - Qincloud' to 'Hongkong - Qincloud 1'	OPTIMIZED in 17.2276056199ms (56.11%)
117	229.279568	254.174107	'Beauharnois - OVH' to 'Hongkong - Qincloud 1'	OPTIMIZED in 24.8945387975ms (9.79%)
118	17.851478	72.304719	'Guanghouz - UCloud' to 'Hongkong - Qincloud 1'	OPTIMIZED in 54.4532411636ms (75.31%)
119	2.133971	1.995446	'Hongkong - Rackspace' to 'Hongkong - Qincloud 1'	OK (in acceptable range 1.99544604317 < 2.1339...
120	57.729732	57.445885	'Tokyo - Amazon' to 'Hongkong - Qincloud 1'	OK (in acceptable range 57.4458848921 < 57.729...
121	160.735741	160.635750	'Strasbourg - OVH' to 'US West - Amazon'	OK (in acceptable range 160.63575 < 160.735741...
122	187.261408	169.855325	'Beijing - Qincloud' to 'US West - Amazon'	OK (in acceptable range 169.855325 < 187.26140...
123	176.315312	175.277662	'Singapour - Amazon' to 'US West - Amazon'	OK (in acceptable range 175.277661871 < 176.31...
124	156.863964	162.888799	'Hongkong - Qincloud 1' to 'US West - Amazon'	OPTIMIZED in 6.02483479304ms (3.70%)
125	160.903761	169.578547	'Hongkong - Qincloud 0' to 'US West - Amazon'	OPTIMIZED in 8.67478589302ms (5.12%)
126	76.815309	83.863021	'Virginia - Amazon' to 'US West - Amazon'	OPTIMIZED in 7.04771207605ms (8.40%)
127	172.103768	284.596410	'Guanghouz - Qincloud' to 'US West - Amazon'	OPTIMIZED in 112.492641956ms (39.53%)
128	70.823345	70.792993	'Beauharnois - OVH' to 'US West - Amazon'	OK (in acceptable range 70.7929928571 < 70.823...
129	174.915174	359.936719	'Guanghouz - UCloud' to 'US West - Amazon'	OPTIMIZED in 185.021545511ms (51.40%)
130	163.996109	157.283007	'Hongkong - Rackspace' to 'US West - Amazon'	OK (in acceptable range 157.283007194 < 163.99...
131	108.594942	105.582734	'Tokyo - Amazon' to 'US West - Amazon'	OK (in acceptable range 105.582733813 < 108.59...

In [30]:

df_packetlost_statistic = pd.concat([go_packetloss_dataframe, hr_packetloss_dataframe, 
                                     bho_packetloss_dataframe, bq_packetloss_dataframe,
                                     gq_packetloss_dataframe, va_packetloss_dataframe, 
                                     hq0_packetloss_dataframe, gu_packetloss_dataframe,
                                     ta_packetloss_dataframe, sa_packetloss_dataframe,
                                     hq1_packetloss_dataframe, wa_packetloss_dataframe
                                    ], ignore_index=True)

df_packetlost_statistic

Out[30]:

	grandenet_ipv6_packet_lost	internet_ipv4_packet_lost	name	state
0	1.147541	2.601626	'Beijing - Qincloud' to 'Strasbourg - OVH'	OPTIMIZED (Better Packet Lost rate)
1	0.000000	0.851064	'Singapour - Amazon' to 'Strasbourg - OVH'	OPTIMIZED (Better Packet Lost rate)
2	0.071429	0.127660	'Hongkong - Qincloud 1' to 'Strasbourg - OVH'	OPTIMIZED (Better Packet Lost rate)
3	0.785714	0.212766	'Hongkong - Qincloud 0' to 'Strasbourg - OVH'	OK (less them 1% diference is considered same)
4	0.215827	0.070922	'US West - Amazon' to 'Strasbourg - OVH'	OK (less them 1% diference is considered same)
5	0.000000	0.000000	'Virginia - Amazon' to 'Strasbourg - OVH'	OK (Same Packet Lost rate)
6	0.214286	18.652482	'Guanghouz - Qincloud' to 'Strasbourg - OVH'	OPTIMIZED (Better Packet Lost rate)
7	0.070922	0.281690	'Beauharnois - OVH' to 'Strasbourg - OVH'	OPTIMIZED (Better Packet Lost rate)
8	0.785714	26.659574	'Guanghouz - UCloud' to 'Strasbourg - OVH'	OPTIMIZED (Better Packet Lost rate)
9	1.357143	1.631206	'Hongkong - Rackspace' to 'Strasbourg - OVH'	OPTIMIZED (Better Packet Lost rate)
10	0.071429	0.283688	'Tokyo - Amazon' to 'Strasbourg - OVH'	OPTIMIZED (Better Packet Lost rate)
11	0.141844	1.549296	'Strasbourg - OVH' to 'Hongkong - Rackspace'	OPTIMIZED (Better Packet Lost rate)
12	0.163934	0.327869	'Beijing - Qincloud' to 'Hongkong - Rackspace'	OPTIMIZED (Better Packet Lost rate)
13	0.000000	0.780142	'Singapour - Amazon' to 'Hongkong - Rackspace'	OPTIMIZED (Better Packet Lost rate)
14	0.000000	0.000000	'Hongkong - Qincloud 1' to 'Hongkong - Rackspace'	OK (Same Packet Lost rate)
15	0.000000	0.000000	'Hongkong - Qincloud 0' to 'Hongkong - Rackspace'	OK (Same Packet Lost rate)
16	2.517986	0.851064	'US West - Amazon' to 'Hongkong - Rackspace'	BAD (Worst Packet Lost rate)
17	1.276596	0.000000	'Virginia - Amazon' to 'Hongkong - Rackspace'	BAD (Worst Packet Lost rate)
18	0.000000	0.141844	'Guanghouz - Qincloud' to 'Hongkong - Rackspace'	OPTIMIZED (Better Packet Lost rate)
19	1.134752	1.760563	'Beauharnois - OVH' to 'Hongkong - Rackspace'	OPTIMIZED (Better Packet Lost rate)
20	0.071429	1.063830	'Guanghouz - UCloud' to 'Hongkong - Rackspace'	OPTIMIZED (Better Packet Lost rate)
21	0.000000	2.198582	'Tokyo - Amazon' to 'Hongkong - Rackspace'	OPTIMIZED (Better Packet Lost rate)
22	0.780142	0.049296	'Strasbourg - OVH' to 'Beauharnois - OVH'	OK (less them 1% diference is considered same)
23	1.074380	0.573770	'Beijing - Qincloud' to 'Beauharnois - OVH'	OK (less them 1% diference is considered same)
24	0.000000	0.212766	'Singapour - Amazon' to 'Beauharnois - OVH'	OPTIMIZED (Better Packet Lost rate)
25	0.000000	0.141844	'Hongkong - Qincloud 1' to 'Beauharnois - OVH'	OPTIMIZED (Better Packet Lost rate)
26	0.000000	0.000000	'Hongkong - Qincloud 0' to 'Beauharnois - OVH'	OK (Same Packet Lost rate)
27	0.143885	0.000000	'US West - Amazon' to 'Beauharnois - OVH'	OK (less them 1% diference is considered same)
28	0.000000	0.000000	'Virginia - Amazon' to 'Beauharnois - OVH'	OK (Same Packet Lost rate)
29	0.214286	10.092199	'Guanghouz - Qincloud' to 'Beauharnois - OVH'	OPTIMIZED (Better Packet Lost rate)
30	1.071429	18.078014	'Guanghouz - UCloud' to 'Beauharnois - OVH'	OPTIMIZED (Better Packet Lost rate)
31	0.857143	1.638298	'Hongkong - Rackspace' to 'Beauharnois - OVH'	OPTIMIZED (Better Packet Lost rate)
32	0.000000	0.141844	'Tokyo - Amazon' to 'Beauharnois - OVH'	OPTIMIZED (Better Packet Lost rate)
33	3.475177	3.450704	'Strasbourg - OVH' to 'Beijing - Qincloud'	OK (less them 1% diference is considered same)
34	1.928571	3.049645	'Singapour - Amazon' to 'Beijing - Qincloud'	OPTIMIZED (Better Packet Lost rate)
35	1.857143	2.127660	'Hongkong - Qincloud 1' to 'Beijing - Qincloud'	OPTIMIZED (Better Packet Lost rate)
36	0.857143	1.560284	'Hongkong - Qincloud 0' to 'Beijing - Qincloud'	OPTIMIZED (Better Packet Lost rate)
37	1.438849	3.404255	'US West - Amazon' to 'Beijing - Qincloud'	OPTIMIZED (Better Packet Lost rate)
38	3.120567	6.478873	'Virginia - Amazon' to 'Beijing - Qincloud'	OPTIMIZED (Better Packet Lost rate)
39	0.000000	0.425532	'Guanghouz - Qincloud' to 'Beijing - Qincloud'	OPTIMIZED (Better Packet Lost rate)
40	1.985816	3.197183	'Beauharnois - OVH' to 'Beijing - Qincloud'	OPTIMIZED (Better Packet Lost rate)
41	0.000000	0.212766	'Guanghouz - UCloud' to 'Beijing - Qincloud'	OPTIMIZED (Better Packet Lost rate)
42	0.142857	0.709220	'Hongkong - Rackspace' to 'Beijing - Qincloud'	OPTIMIZED (Better Packet Lost rate)
43	1.357143	0.765957	'Tokyo - Amazon' to 'Beijing - Qincloud'	OK (less them 1% diference is considered same)
44	1.347518	24.443662	'Strasbourg - OVH' to 'Guanghouz - Qincloud'	OPTIMIZED (Better Packet Lost rate)
45	0.000000	0.000000	'Beijing - Qincloud' to 'Guanghouz - Qincloud'	OK (Same Packet Lost rate)
46	0.214286	0.212766	'Singapour - Amazon' to 'Guanghouz - Qincloud'	OK (less them 1% diference is considered same)
47	0.214286	1.134752	'Hongkong - Qincloud 1' to 'Guanghouz - Qincloud'	OPTIMIZED (Better Packet Lost rate)
48	0.642857	1.134752	'Hongkong - Qincloud 0' to 'Guanghouz - Qincloud'	OPTIMIZED (Better Packet Lost rate)
49	2.086331	14.014184	'US West - Amazon' to 'Guanghouz - Qincloud'	OPTIMIZED (Better Packet Lost rate)
50	1.134752	10.464789	'Virginia - Amazon' to 'Guanghouz - Qincloud'	OPTIMIZED (Better Packet Lost rate)
51	0.354610	22.147887	'Beauharnois - OVH' to 'Guanghouz - Qincloud'	OPTIMIZED (Better Packet Lost rate)
52	0.142857	0.070922	'Guanghouz - UCloud' to 'Guanghouz - Qincloud'	OK (less them 1% diference is considered same)
53	0.714286	0.000000	'Hongkong - Rackspace' to 'Guanghouz - Qincloud'	OK (less them 1% diference is considered same)
54	1.000000	8.865248	'Tokyo - Amazon' to 'Guanghouz - Qincloud'	OPTIMIZED (Better Packet Lost rate)
55	0.070922	0.000000	'Strasbourg - OVH' to 'Virginia - Amazon'	OK (less them 1% diference is considered same)
56	0.491803	0.975610	'Beijing - Qincloud' to 'Virginia - Amazon'	OPTIMIZED (Better Packet Lost rate)
57	0.000000	0.070922	'Singapour - Amazon' to 'Virginia - Amazon'	OPTIMIZED (Better Packet Lost rate)
58	0.000000	2.553191	'Hongkong - Qincloud 1' to 'Virginia - Amazon'	OPTIMIZED (Better Packet Lost rate)
59	0.000000	0.000000	'Hongkong - Qincloud 0' to 'Virginia - Amazon'	OK (Same Packet Lost rate)
60	0.000000	0.000000	'US West - Amazon' to 'Virginia - Amazon'	OK (Same Packet Lost rate)
61	0.071429	0.992908	'Guanghouz - Qincloud' to 'Virginia - Amazon'	OPTIMIZED (Better Packet Lost rate)
62	0.000000	0.000000	'Beauharnois - OVH' to 'Virginia - Amazon'	OK (Same Packet Lost rate)
63	0.357143	21.652482	'Guanghouz - UCloud' to 'Virginia - Amazon'	OPTIMIZED (Better Packet Lost rate)
64	0.571429	0.070922	'Hongkong - Rackspace' to 'Virginia - Amazon'	OK (less them 1% diference is considered same)
65	0.000000	0.000000	'Tokyo - Amazon' to 'Virginia - Amazon'	OK (Same Packet Lost rate)
66	1.631206	0.774648	'Strasbourg - OVH' to 'Hongkong - Qincloud 0'	OK (less them 1% diference is considered same)
67	0.330579	0.983607	'Beijing - Qincloud' to 'Hongkong - Qincloud 0'	OPTIMIZED (Better Packet Lost rate)
68	0.000000	0.000000	'Singapour - Amazon' to 'Hongkong - Qincloud 0'	OK (Same Packet Lost rate)
69	0.000000	0.000000	'Hongkong - Qincloud 1' to 'Hongkong - Qinclou...	OK (Same Packet Lost rate)
70	0.575540	0.000000	'US West - Amazon' to 'Hongkong - Qincloud 0'	OK (less them 1% diference is considered same)
71	0.000000	0.000000	'Virginia - Amazon' to 'Hongkong - Qincloud 0'	OK (Same Packet Lost rate)
72	0.857143	0.851064	'Guanghouz - Qincloud' to 'Hongkong - Qincloud 0'	OK (less them 1% diference is considered same)
73	0.000000	0.492958	'Beauharnois - OVH' to 'Hongkong - Qincloud 0'	OPTIMIZED (Better Packet Lost rate)
74	0.142857	4.680851	'Guanghouz - UCloud' to 'Hongkong - Qincloud 0'	OPTIMIZED (Better Packet Lost rate)
75	0.000000	0.000000	'Hongkong - Rackspace' to 'Hongkong - Qincloud 0'	OK (Same Packet Lost rate)
76	0.000000	0.141844	'Tokyo - Amazon' to 'Hongkong - Qincloud 0'	OPTIMIZED (Better Packet Lost rate)
77	0.709220	17.605634	'Strasbourg - OVH' to 'Guanghouz - UCloud'	OPTIMIZED (Better Packet Lost rate)
78	0.000000	0.409836	'Beijing - Qincloud' to 'Guanghouz - UCloud'	OPTIMIZED (Better Packet Lost rate)
79	0.785714	0.709220	'Singapour - Amazon' to 'Guanghouz - UCloud'	OK (less them 1% diference is considered same)
80	0.357143	2.808511	'Hongkong - Qincloud 1' to 'Guanghouz - UCloud'	OPTIMIZED (Better Packet Lost rate)
81	0.071429	4.326241	'Hongkong - Qincloud 0' to 'Guanghouz - UCloud'	OPTIMIZED (Better Packet Lost rate)
82	3.237410	21.588652	'US West - Amazon' to 'Guanghouz - UCloud'	OPTIMIZED (Better Packet Lost rate)
83	0.780142	23.556338	'Virginia - Amazon' to 'Guanghouz - UCloud'	OPTIMIZED (Better Packet Lost rate)
84	0.000000	0.000000	'Guanghouz - Qincloud' to 'Guanghouz - UCloud'	OK (Same Packet Lost rate)
85	3.191489	20.669014	'Beauharnois - OVH' to 'Guanghouz - UCloud'	OPTIMIZED (Better Packet Lost rate)
86	0.928571	1.560284	'Hongkong - Rackspace' to 'Guanghouz - UCloud'	OPTIMIZED (Better Packet Lost rate)
87	0.857143	8.297872	'Tokyo - Amazon' to 'Guanghouz - UCloud'	OPTIMIZED (Better Packet Lost rate)
88	0.070922	0.352113	'Strasbourg - OVH' to 'Tokyo - Amazon'	OPTIMIZED (Better Packet Lost rate)
89	0.826446	0.327869	'Beijing - Qincloud' to 'Tokyo - Amazon'	OK (less them 1% diference is considered same)
90	0.000000	0.000000	'Singapour - Amazon' to 'Tokyo - Amazon'	OK (Same Packet Lost rate)
91	0.000000	0.000000	'Hongkong - Qincloud 1' to 'Tokyo - Amazon'	OK (Same Packet Lost rate)
92	0.142857	0.141844	'Hongkong - Qincloud 0' to 'Tokyo - Amazon'	OK (less them 1% diference is considered same)
93	0.000000	0.000000	'US West - Amazon' to 'Tokyo - Amazon'	OK (Same Packet Lost rate)
94	0.000000	0.000000	'Virginia - Amazon' to 'Tokyo - Amazon'	OK (Same Packet Lost rate)
95	0.214286	10.567376	'Guanghouz - Qincloud' to 'Tokyo - Amazon'	OPTIMIZED (Better Packet Lost rate)
96	0.070922	0.211268	'Beauharnois - OVH' to 'Tokyo - Amazon'	OPTIMIZED (Better Packet Lost rate)
97	0.785714	9.716312	'Guanghouz - UCloud' to 'Tokyo - Amazon'	OPTIMIZED (Better Packet Lost rate)
98	0.000000	2.553191	'Hongkong - Rackspace' to 'Tokyo - Amazon'	OPTIMIZED (Better Packet Lost rate)
99	0.212766	0.985915	'Strasbourg - OVH' to 'Singapour - Amazon'	OPTIMIZED (Better Packet Lost rate)
100	0.245902	1.056911	'Beijing - Qincloud' to 'Singapour - Amazon'	OPTIMIZED (Better Packet Lost rate)
101	0.000000	0.000000	'Hongkong - Qincloud 1' to 'Singapour - Amazon'	OK (Same Packet Lost rate)
102	0.000000	0.000000	'Hongkong - Qincloud 0' to 'Singapour - Amazon'	OK (Same Packet Lost rate)
103	0.000000	0.000000	'US West - Amazon' to 'Singapour - Amazon'	OK (Same Packet Lost rate)
104	0.000000	0.000000	'Virginia - Amazon' to 'Singapour - Amazon'	OK (Same Packet Lost rate)
105	0.214286	0.141844	'Guanghouz - Qincloud' to 'Singapour - Amazon'	OK (less them 1% diference is considered same)
106	0.070922	0.140845	'Beauharnois - OVH' to 'Singapour - Amazon'	OPTIMIZED (Better Packet Lost rate)
107	0.071429	0.638298	'Guanghouz - UCloud' to 'Singapour - Amazon'	OPTIMIZED (Better Packet Lost rate)
108	0.000000	1.347518	'Hongkong - Rackspace' to 'Singapour - Amazon'	OPTIMIZED (Better Packet Lost rate)
109	0.000000	0.000000	'Tokyo - Amazon' to 'Singapour - Amazon'	OK (Same Packet Lost rate)
110	1.134752	0.492958	'Strasbourg - OVH' to 'Hongkong - Qincloud 1'	OK (less them 1% diference is considered same)
111	0.991736	1.229508	'Beijing - Qincloud' to 'Hongkong - Qincloud 1'	OPTIMIZED (Better Packet Lost rate)
112	0.071429	0.000000	'Singapour - Amazon' to 'Hongkong - Qincloud 1'	OK (less them 1% diference is considered same)
113	0.000000	0.000000	'Hongkong - Qincloud 0' to 'Hongkong - Qinclou...	OK (Same Packet Lost rate)
114	0.000000	0.000000	'US West - Amazon' to 'Hongkong - Qincloud 1'	OK (Same Packet Lost rate)
115	0.000000	2.957746	'Virginia - Amazon' to 'Hongkong - Qincloud 1'	OPTIMIZED (Better Packet Lost rate)
116	0.285714	0.780142	'Guanghouz - Qincloud' to 'Hongkong - Qincloud 1'	OPTIMIZED (Better Packet Lost rate)
117	0.070922	0.633803	'Beauharnois - OVH' to 'Hongkong - Qincloud 1'	OPTIMIZED (Better Packet Lost rate)
118	0.000000	1.631206	'Guanghouz - UCloud' to 'Hongkong - Qincloud 1'	OPTIMIZED (Better Packet Lost rate)
119	0.000000	0.070922	'Hongkong - Rackspace' to 'Hongkong - Qincloud 1'	OPTIMIZED (Better Packet Lost rate)
120	0.000000	0.070922	'Tokyo - Amazon' to 'Hongkong - Qincloud 1'	OPTIMIZED (Better Packet Lost rate)
121	0.141844	0.000000	'Strasbourg - OVH' to 'US West - Amazon'	OK (less them 1% diference is considered same)
122	0.491803	0.327869	'Beijing - Qincloud' to 'US West - Amazon'	OK (less them 1% diference is considered same)
123	0.000000	0.000000	'Singapour - Amazon' to 'US West - Amazon'	OK (Same Packet Lost rate)
124	0.000000	0.000000	'Hongkong - Qincloud 1' to 'US West - Amazon'	OK (Same Packet Lost rate)
125	0.928571	0.000000	'Hongkong - Qincloud 0' to 'US West - Amazon'	OK (less them 1% diference is considered same)
126	0.070922	0.000000	'Virginia - Amazon' to 'US West - Amazon'	OK (less them 1% diference is considered same)
127	1.642857	16.439716	'Guanghouz - Qincloud' to 'US West - Amazon'	OPTIMIZED (Better Packet Lost rate)
128	0.000000	0.140845	'Beauharnois - OVH' to 'US West - Amazon'	OPTIMIZED (Better Packet Lost rate)
129	1.785714	23.531915	'Guanghouz - UCloud' to 'US West - Amazon'	OPTIMIZED (Better Packet Lost rate)
130	3.357143	0.921986	'Hongkong - Rackspace' to 'US West - Amazon'	BAD (Worst Packet Lost rate)
131	0.000000	0.000000	'Tokyo - Amazon' to 'US West - Amazon'	OK (Same Packet Lost rate)

我们可以有多快？是否有速度限制？

我们监测到的实验结果超出了我们的最高设想。我们发现与标准 IPv4 相比，GrandeNet 整体降低网络延迟15%到30%, 数据流失降到少于1% (5-10倍的提升)。

考虑到被监测的服务器之间的所有独立链接和 ping 响应时间 (网络延迟）, 50%和70%的连接都通过 GrandeNet 优化了，45%到65%的连接数据流失都得到有效降低。

由于 GrandeNet 还没有使用一个完整的网格配置，所以我们同时还观察到使用 GrandeNet IPv6 时，一小部分连接的质量下降。但是这个问题可以通过创建额外的服务器之间的直接通道来解决，监测结果表明，它要比标准 IPv4 更快。通过创建直接通道，我们可以稳定网格质量，因为它只包含优化了的或 GrandeNet 上有同样速度的链接 (额外的直接通道)。

In [31]:

# Global Average result

ipv4_mean = df_average_statistic["internet_ipv4"].mean() 
ipv6_mean = df_average_statistic["grandenet_ipv6"].mean()

display("IPv4 Final Average from all links: %.2fms" % ipv4_mean)
display("IPv6 Final Average from all links: %.2fms" % ipv6_mean)

if ipv4_mean > ipv6_mean:
    display("Grandent Optimized the network Globally in %.2f%% in average for each ping response." % \
            (((ipv4_mean - ipv6_mean)/ipv4_mean)*100))
elif ipv4_mean == ipv6_mean:
    display("<h1>No Optimisation found.</h1>")
elif ipv4_mean < ipv6_mean:
    display("Grandenet is making things worst in %sms, please review the configuration." % (ipv6_mean - ipv4_mean))

'IPv4 Final Average from all links: 161.75ms'

'IPv6 Final Average from all links: 127.46ms'

'Grandent Optimized the network Globally in 21.20% in average for each ping response.'

In [32]:

# Number of optimisations
total = len(df_average_statistic)
optimized = len(df_average_statistic[df_average_statistic["internet_ipv4"] > df_average_statistic["grandenet_ipv6"]])

worst = len(df_average_statistic[
    df_average_statistic.apply(lambda x: x["internet_ipv4"] < x["grandenet_ipv6"] - max(20, x["internet_ipv4"]*0.1), axis=1)])


display("Total Number of Links tested: %s" % total)
display("Number of Optimized Links: %s (%.2f%%)" % (optimized, (float(optimized)/total)*100))
display("Number of links which becomes worst: %s (%.2f%%)" % (worst, (float(worst)/total)*100))
display("Number of links which unchanged : %s (%.2f%%)" % ((total-(optimized+worst)), (float(total-(optimized+worst))/total)*100))

'Total Number of Links tested: 132'

'Number of Optimized Links: 78 (59.09%)'

'Number of links which becomes worst: 0 (0.00%)'

'Number of links which unchanged : 54 (40.91%)'

In [33]:

ipv4_package_lost_rate = df_packetlost_statistic["internet_ipv4_packet_lost"].mean()
ipv6_package_lost_rate = df_packetlost_statistic["grandenet_ipv6_packet_lost"].mean()

display("Average of Packet Lost Rate per link using Internet (IPv4): %.2f%%" % ipv4_package_lost_rate)
display("Average of Packet Lost Rate per link using Grandenet (IPv6): %.2f%%" % ipv6_package_lost_rate)


if ipv4_package_lost_rate > ipv6_package_lost_rate:
    display("Grandent Optimized the network Globally in %.2f%% less packet lost in average, it is %.2f times better." % \
            (ipv4_package_lost_rate - ipv6_package_lost_rate, (float(ipv4_package_lost_rate/ipv6_package_lost_rate))))
elif ipv4_package_lost_rate == ipv6_package_lost_rate:
    display("No Optimisation found.")
elif ipv4_package_lost_rate < ipv6_package_lost_rate:
    display("Grandenet is making things worst in %.2f%% , please review the configuration." % (ipv6_package_lost_rate - ipv4_package_lost_rate))

'Average of Packet Lost Rate per link using Internet (IPv4): 3.12%'

'Average of Packet Lost Rate per link using Grandenet (IPv6): 0.52%'

'Grandent Optimized the network Globally in 2.61% less packet lost in average, it is 6.05 times better.'

In [34]:

dps = df_packetlost_statistic

total = len(dps)
optimized = len(dps[dps["internet_ipv4_packet_lost"] > dps["grandenet_ipv6_packet_lost"]])
worst = len(dps[(dps["internet_ipv4_packet_lost"] < dps["grandenet_ipv6_packet_lost"])]
                [(dps["grandenet_ipv6_packet_lost"] - dps["internet_ipv4_packet_lost"]) > 1])
same = len(dps[(dps["internet_ipv4_packet_lost"] == dps["grandenet_ipv6_packet_lost"])])

same += len(dps[(dps["grandenet_ipv6_packet_lost"] - dps["internet_ipv4_packet_lost"]) < 1]
                 [(dps["grandenet_ipv6_packet_lost"] > dps["internet_ipv4_packet_lost"])])

display("Total Number of Links tested: %s" % total)
display("Number of Optimized links with less packet lost: %s (%.2f%%)" % (optimized, (float(optimized)/total)*100))
display("Number of links which becomes worst: %s (%.2f%%)" % (worst, (float(worst)/total)*100))
display("Number of links which unchanged : %s (%.2f%%)" % (same, (float(same)/total)*100))

'Total Number of Links tested: 132'

'Number of Optimized links with less packet lost: 75 (56.82%)'

'Number of links which becomes worst: 3 (2.27%)'

'Number of links which unchanged : 54 (40.91%)'

总结

根据以上例子我们可以发现，相比标准 IPv4 连接而言，GrandeNet IPv6 降低了15-30%的网络延迟，数据流失降到低于1% (5-10倍的提升）。50-70%的连接可以被优化，45-65%的连接可以减少数据丢失。GrandeNet IPv6 最终的平均响应时间127.46毫秒，远低于默认的IPv4平均响应时间161.75毫秒。

使用 GrandeNet，整体21.20%的 pings 都被优化了，数据丢失从3.12% (IPv4) 降至0.52 (IPv6), 说明监测和网络连接优化可以带来显著的效益。创建工具自动化了我们的实验，并为用户提供了易于使用的应用，实时展示路径情况以及连接问题将是下一个研究方向。

GrandeNet - 互联网加速剂