VPS推荐网搬瓦工在双十一期间推出的一款特价KVM VPS主机方案货源已经补充过几次了，目前均还有货源，小七在第一时间入手了一台，基本上处于压箱子的状态，依然是默认的洛杉矶QN（DC3）机房。在这个折腾比较多的时间段，简单看看其表现吧。
一、配置方案：
- 核心：2核CPU
- 内存：2GB
- 硬盘：40GB
- 流量：1TB
- 端口：1Gbps
- 架构：KVM
- 优惠码：BWH26FXH3HIQ 可以节约一点点；
- 价格：$29.88/年
- 传送：购买链接
注：在执行wget请求的时候出现wget: command not found这样的一个提示，其实是系统WEGT 没有安装。安装的方式非常简单，只需要以下的一条命令就行啦！
输入命令：yum -y install wget 就会自动安装WGET，这样问题就解决了！
二、基本信息：
Benchmark started on 15-Nov-2018 02:41:35
## System Information
OS Name : CentOS release 6.8 (32 bit)
Kernel : KVM / 4.10.4-1.el6.elrepo.i686
Hostname : xqblog.com
CPU Model : QEMU Virtual CPU version (cpu64-rhel6)
CPU Cores : 2 cores @ 2599.998 MHz
CPU Cache : 4096 KB
Total RAM : 2030 MiB (Free 92 MiB)
Total SWAP : 511 MiB (Free 511MiB)
Total Space : 40GB (4% used)
Running for : 9 minutes 10 seconds
## CDN Speedtest
CacheFly : 78.68 MiB/s | 629.46 Mbps | ping 0.493ms
CDN.net : 824 B/s | N/A | ping 0.581ms
Gdrive : 0 B/s | N/A | ping 0.384ms
## North America Speedtest
Softlayer, Washington, USA : 20.32 MiB/s | 162.58 Mbps | ping 61.655ms
SoftLayer, San Jose, USA : 68.97 MiB/s | 551.79 Mbps | ping 9.851ms
SoftLayer, Dallas, USA : 0 B/s | N/A | ping error!
Vultr, New Jersey, USA : 31.92 MiB/s | 255.36 Mbps | ping 63.150ms
Vultr, Seattle, USA : 66.63 MiB/s | 533.08 Mbps | ping 25.901ms
Vultr, Dallas, USA : 56.24 MiB/s | 449.95 Mbps | ping 34.308ms
Vultr, Los Angeles, USA : 11.22 MiB/s | 89.73 Mbps | ping 0.670ms
Ramnode, New York, USA : 20.30 MiB/s | 162.44 Mbps | ping 62.790ms
Ramnode, Atlanta, USA : 25.16 MiB/s | 201.28 Mbps | ping 45.046ms
OVH, Beauharnois, Canada : 9.87 MiB/s | 78.94 Mbps | ping 68.718ms
## Europe Speedtest
Vultr, London, UK : 12.20 MiB/s | 97.64 Mbps | ping 132.922ms
LeaseWeb, Frankfurt, Germany : 7.58 MiB/s | 60.67 Mbps | ping 151.174ms
Hetzner, Germany : 5.86 MiB/s | 46.91 Mbps | ping 149.222ms
Ramnode, Alblasserdam, NL : 1.03 MiB/s | 8.26 Mbps | ping 141.269ms
Vultr, Amsterdam, NL : 12.93 MiB/s | 103.42 Mbps | ping 151.591ms
EDIS, Stockholm, Sweden : 10.48 MiB/s | 83.87 Mbps | ping 159.202ms
OVH, Roubaix, France : 5.97 MiB/s | 47.80 Mbps | ping 135.894ms
Online, France : 3.98 MiB/s | 31.85 Mbps | ping 137.717ms
Prometeus, Milan, Italy : 12.16 MiB/s | 97.25 Mbps | ping 149.932ms
## Exotic Speedtest
Sydney, Australia : 7.46 MiB/s | 59.69 Mbps | ping 159.522ms
Lagoon, New Caledonia : 406.08 KiB/s | 3.17 Mbps | ping 170.118ms
Hosteasy, Moldova : 212.75 KiB/s | 1.66 Mbps | ping 189.133ms
Prima, Argentina : 620.85 KiB/s | 4.85 Mbps | ping error!
## Asia Speedtest
SoftLayer, Singapore : 685.49 KiB/s | 5.36 Mbps | ping 174.153ms
Linode, Tokyo, Japan : 6.79 MiB/s | 54.29 Mbps | ping 163.161ms
Linode, Singapore : 1.18 MiB/s | 9.47 Mbps | ping 188.310ms
Vultr, Tokyo, Japan : 18.73 MiB/s | 149.85 Mbps | ping 115.497ms
## IO Test
CPU Speed:
bzip2 512MB – 48.9 MB/s
sha256 512MB – 98.6 MB/s
md5sum 512MB – 270 MB/s
Disk Speed (512MB):
I/O Speed – 240 MB/s
I/O Direct – 110 MB/s
RAM Speed (1024MB):
Avg. write – 877.9 MB/s
Avg. read – 2048.0 MB/s
另外脚本测试基本信息：
-------------------------------------------------------------------------- CPU model : QEMU Virtual CPU version (cpu64-rhel6) Number of cores : 2 CPU frequency : 2599.998 MHz Total size of Disk : 40.3 GB (1.4 GB Used) Total amount of Mem : 2030 MB (178 MB Used) Total amount of Swap : 511 MB (0 MB Used) System uptime : 0 days, 0 hour 14 min Load average : 0.12, 0.10, 0.04 OS : CentOS 6.8 Arch : i686 (32 Bit) Kernel : 4.10.4-1.el6.elrepo.i686 Virt : kvm -------------------------------------------------------------------------- I/O speed(1st run) :517 MB/s I/O speed(2nd run) :358 MB/s I/O speed(3rd run) :502 MB/s -------------------------------------------------------------------------- Node Name IP Address Download Speed Latency CacheFly 205.234.175.175 87.8MB/s 0.515 ms Linode, Tokyo, JP 106.187.96.148 4.17MB/s 163.151 ms Linode, Singapore, SG 139.162.23.4 8.32MB/s 188.355 ms Linode, London, UK 176.58.107.39 11.5MB/s 135.087 ms Linode, Frankfurt, DE 139.162.130.8 7.82MB/s 156.829 ms Linode, Fremont, CA 50.116.14.9 13.9MB/s 61.421 ms Softlayer, Dallas, TX 173.192.68.18 40.0MB/s 30.502 ms Softlayer, Seattle, WA 67.228.112.250 44.5MB/s 28.442 ms Softlayer, Frankfurt, DE 159.122.69.4 5.55MB/s 145.162 ms Softlayer, Singapore, SG 119.81.28.170 7.67MB/s 174.122 ms Softlayer, HongKong, CN 119.81.130.170 8.71MB/s 168.402 ms -------------------------------------------------------------------------- Node Name Upload Speed Download Speed Latency Xiangyang CT 75.55 Mbit/s 41.87 Mbit/s 253.133 ms Xi'an CU 9.82 Mbit/s 80.87 Mbit/s 225.969 ms Shanghai CU 126.14 Mbit/s 288.38 Mbit/s 188.872 ms Chongqing CU 81.46 Mbit/s 34.79 Mbit/s 186.317 ms Hunan CT 16.18 Mbit/s 11.26 Mbit/s 190.582 ms -------------------------------------------------------------------------- Zhengzhou : 188.9 ms Shenyang : 169.17 ms Tianjin : 167.76 ms Changji : Fail Lhasa : 223.27 ms Hefei : 172.18 ms Shenzhen : Fail Lanzhou : 219.25 ms Xining : 206.88 ms Changchun : 181.7 ms Wuhan : 164.92 ms Xiangyang : Fail Nanchang : Fail Chongqing : 186.52 ms Shanghai : 132.99 ms Huhehaote : 200.59 ms Urumqi : 218.49 ms Hangzhou : 130.8 ms Xi an : 188.99 ms Ningbo : 150.21 ms Taiyuan : Fail Suzhou : 139.92 ms Changsha : 200.15 ms Harbin : 193.56 ms Beijing : 156.54 ms Chengdu : 176.98 ms Jinan : 171.14 ms Nanjing : 147.09 ms Ningxia : 199.31 ms Fuzhou : 178.49 ms --------------------------------------------------------------------------
注：两个脚本测试还是有一点点差距的；
三、硬盘IO性能
I/O speed(1st run) :517 MB/s I/O speed(2nd run) :358 MB/s I/O speed(3rd run) :502 MB/s
四、国内上传与下载：
Node Name Upload Speed Download Speed Latency 襄阳电信 1.52 Mbit/s 0.00 Mbit/s 000.000 ms 昆明电信 112.26 Mbit/s 39.71 Mbit/s 154.971 ms 沈阳联通 108.74 Mbit/s 275.79 Mbit/s 249.566 ms 西安联通 58.02 Mbit/s 82.99 Mbit/s 211.882 ms 上海联通 111.19 Mbit/s 299.06 Mbit/s 173.0 ms 重庆联通 82.53 Mbit/s 32.31 Mbit/s 185.891 ms 西安移动 71.62 Mbit/s 18.04 Mbit/s 209.864 ms 宁波移动 8.38 Mbit/s 13.35 Mbit/s 872.047 ms 成都移动 15.33 Mbit/s 12.47 Mbit/s 209.139 ms 花费时间：4 分 7 秒 当前时间: 2018-11-15 03:00:58 全面测试完成！
五、UnixBench跑分
------------------------------------------------------------------------
Benchmark Run: Thu Nov 15 2018 03:36:58 - 04:04:51
2 CPUs in system; running 2 parallel copies of tests
Dhrystone 2 using register variables 35006481.5 lps (10.0 s, 7 samples)
Double-Precision Whetstone 4408.2 MWIPS (10.1 s, 7 samples)
Execl Throughput 5304.9 lps (29.9 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks 320693.5 KBps (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks 85504.9 KBps (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks 1052429.5 KBps (30.0 s, 2 samples)
Pipe Throughput 576839.8 lps (10.0 s, 7 samples)
Pipe-based Context Switching 156906.8 lps (10.0 s, 7 samples)
Process Creation 7775.7 lps (30.0 s, 2 samples)
Shell Scripts (1 concurrent) 5889.4 lpm (60.0 s, 2 samples)
Shell Scripts (8 concurrent) 787.2 lpm (60.1 s, 2 samples)
System Call Overhead 2097095.7 lps (10.0 s, 7 samples)
System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 35006481.5 2999.7
Double-Precision Whetstone 55.0 4408.2 801.5
Execl Throughput 43.0 5304.9 1233.7
File Copy 1024 bufsize 2000 maxblocks 3960.0 320693.5 809.8
File Copy 256 bufsize 500 maxblocks 1655.0 85504.9 516.6
File Copy 4096 bufsize 8000 maxblocks 5800.0 1052429.5 1814.5
Pipe Throughput 12440.0 576839.8 463.7
Pipe-based Context Switching 4000.0 156906.8 392.3
Process Creation 126.0 7775.7 617.1
Shell Scripts (1 concurrent) 42.4 5889.4 1389.0
Shell Scripts (8 concurrent) 6.0 787.2 1312.0
System Call Overhead 15000.0 2097095.7 1398.1
========
System Benchmarks Index Score 964.0
六、小结
目前宿主主机几乎处于饱和状态，跑分花了不少的时间，整体上来看只能算是入门的水平，不过用于搭建一个小站或者网络加速服务还是可以的，稳定性以及网络质量方面估计要等段时间，等折腾的少了之后也许会好一些。现在无法切换的洛杉矶C3（DC8）机房，洛杉矶QN（DC3）默认购买，追求稳定一些的用户可以切换到其他节点，比如洛杉矶MC（非优化线路），晚点小七切换到MC节点看看其性能如何吧。