0x01 前言
前些天我测试了krypt的独立服务器，这两天经过与krypt的客服沟通，获得一台新加坡云服务器的48小时测试时间。但经过测试，整体性能和网络状况也并不是很好。
不知道是不是因为是demo主机的原因，磁盘性能依旧不容乐观，相比我之前测试的一些主机，IO确实不太理想。至于网络，我并没有选择中国优化网络，因为新加坡机房的中国优化网络需要25美元，加上服务器租用费用，总月付费用高达50美元。
50美元/月，我已经有很多很好的选择了，以下是测试结果。
0x02 配置
母鸡CPU是E5-2620 v2，云主机配置如下：
- CPU：1核心
- 内存：1GB
- 网卡：千兆
- 端口速度：20Mbps
- 硬盘：30G
这个服务器并没有控制面板，因为这是demo机子，需要通过客服开通。如果自行购买的话，费用如下：
选择中国优化网络会走CN2线路，每月需要加25美元。另外需要关注的是，不同型号的端口速度也是不同的。
0x03 网络测试
0x03.1 ping
--- 43.252.120.61 ping statistics --- 4196 packets transmitted, 4052 received, 3% packet loss, time 4197461ms rtt min/avg/max/mdev = 204.874/246.467/527.193/18.073 ms
虽然新加坡比美国要近，可是延迟却比美国要差，主要是线路不太好，很绕。
0x03.2 wget
以下是深圳电信通过wget下载服务器上的文件：
[root@misc ~]# wget http://43.252.120.61/1gb-1.test --2017-07-06 21:42:06-- http://43.252.120.61/1gb-1.test Connecting to 43.252.120.61:80... connected. HTTP request sent, awaiting response... 200 OK Length: 1073741824 (1.0G) [application/octet-stream] Saving to: ‘1gb-1.test.1’ 0% [ ] 171,095 5.25KB/s eta 5d 12h
速度非常缓慢，很明显被电信QOS限速了。而深圳联通和移动的速度却高达2m/s。
以下是香港国际网络wget服务器上文件的结果：
[root@hk1 ~]$ wget http://43.252.120.61/1gb-1.test --2017-07-06 21:42:34-- http://43.252.120.61/1gb-1.test Connecting to 43.252.120.61:80... connected. HTTP request sent, awaiting response... 200 OK Length: 1073741824 (1.0G) [application/octet-stream] Saving to: ‘1gb-1.test’ 5% [============> ] 62,273,115 2.34MB/s eta 6m 54s
下载速度可以达到服务器的端口速度，这说明网络是挺好的，也可以证明电信对这个服务器的IP做了限速。从客服那里了解到，CN2的速度和延迟都要比国际网络要好，毕竟是电信自己家的网络。
0x03.3 路由追踪
以下是路由追踪的结果：
线路不太好，数据包环游世界。如果资金允许，推荐购买CN2线路，但是要注意端口速度。
0x03.4 speedtest
以下是speedtest的测试结果：
#天津联通 Retrieving speedtest.net configuration... Testing from VPLS (43.252.120.61)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by China Unicom (Tianjin) [3263.87 km]: 628.367 ms Testing download speed................................................................................ Download: 0.91 Mbit/s Testing upload speed...................................................................................................... Upload: 9.03 Mbit/s #青岛联通 Retrieving speedtest.net configuration... Testing from VPLS (43.252.120.61)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by China Unicom (Qingdao) [3178.11 km]: 130.191 ms Testing download speed................................................................................ Download: 18.25 Mbit/s Testing upload speed...................................................................................................... Upload: 18.06 Mbit/s #上海电信 Retrieving speedtest.net configuration... Testing from VPLS (43.252.120.61)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by China Telecom (Shanghai) [2893.58 km]: 805.356 ms Testing download speed................................................................................ Download: 1.08 Mbit/s Testing upload speed...................................................................................................... Upload: 1.05 Mbit/s #浙江移动 Retrieving speedtest.net configuration... Testing from VPLS (43.252.120.61)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by China Mobile Group Zhejiang Branch (Hangzhou) [2728.15 km]: 240.606 ms Testing download speed................................................................................ Download: 13.61 Mbit/s Testing upload speed...................................................................................................... Upload: 14.88 Mbit/s #USTC Retrieving speedtest.net configuration... Testing from VPLS (43.252.120.61)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by University of Electronic Science and Technology of China (Chengdu) [1915.90 km]: 484.283 ms Testing download speed................................................................................ Download: 15.35 Mbit/s Testing upload speed...................................................................................................... Upload: 2.26 Mbit/s #深圳移动 Retrieving speedtest.net configuration... Testing from VPLS (43.252.120.61)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by China Mobile Group GuangDong (Shenzhen) [1741.54 km]: 103.381 ms Testing download speed................................................................................ Download: 5.73 Mbit/s Testing upload speed...................................................................................................... Upload: 19.61 Mbit/s #香港 Retrieving speedtest.net configuration... Testing from VPLS (43.252.120.61)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by Interoute VDC (Hong Kong) [1728.10 km]: 264.189 ms Testing download speed................................................................................ Download: 18.93 Mbit/s Testing upload speed...................................................................................................... Upload: 19.99 Mbit/s #昆明联通 Retrieving speedtest.net configuration... Testing from VPLS (43.252.120.61)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by Yunnan Chinaunicom (Kunming) [1279.56 km]: 212.803 ms Testing download speed................................................................................ Download: 18.00 Mbit/s Testing upload speed...................................................................................................... Upload: 13.04 Mbit/s #昆明电信 Retrieving speedtest.net configuration... Testing from VPLS (43.252.120.61)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by China Telecom (Kunming) [1279.56 km]: 837.997 ms Testing download speed................................................................................ Download: 0.48 Mbit/s Testing upload speed...................................................................................................... Upload: 0.41 Mbit/s
联通和移动的速度都非常好，而电信真是无话可说了。
0x04 性能测试
0x04.1 dd磁盘测试
[root@CloudDemo-9904 ~]# dd bs=1M count=1024 if=/dev/zero of=1gb-1.test conv=fdatasync 1024+0 records in 1024+0 records out 1073741824 bytes (1.1 GB) copied, 14.7637 s, 72.7 MB/s [root@CloudDemo-9904 ~]# dd bs=1M count=1024 if=/dev/zero of=1gb-2.test conv=fdatasync 1024+0 records in 1024+0 records out 1073741824 bytes (1.1 GB) copied, 19.6075 s, 54.8 MB/s [root@CloudDemo-9904 ~]# dd bs=1M count=1024 if=/dev/zero of=1gb-3.test conv=fdatasync 1024+0 records in 1024+0 records out 1073741824 bytes (1.1 GB) copied, 15.977 s, 67.2 MB/s [root@CloudDemo-9904 ~]# dd bs=1M count=1024 if=/dev/zero of=1gb-4.test conv=fdatasync 1024+0 records in 1024+0 records out 1073741824 bytes (1.1 GB) copied, 27.4817 s, 39.1 MB/s [root@CloudDemo-9904 ~]# dd bs=1M count=1024 if=/dev/zero of=1gb-5.test conv=fdatasync 1024+0 records in 1024+0 records out 1073741824 bytes (1.1 GB) copied, 22.2669 s, 48.2 MB/s
很明显用的是机械硬盘，而且SSD缓存都没上，速度非常慢。
0x04.2 fio
CRC 加密测试：
[root@CloudDemo-9904 fio-master]# ./fio -crctest md5: 351.14 MiB/sec crc64: 270.47 MiB/sec crc32: 233.57 MiB/sec crc32c: 5927.02 MiB/sec crc16: 270.98 MiB/sec crc7: 246.62 MiB/sec sha1: 311.31 MiB/sec sha256: 103.97 MiB/sec sha512: 184.86 MiB/sec xxhash: 4066.01 MiB/sec murmur3: 1764.51 MiB/sec jhash: 1627.48 MiB/sec fnv: 4454.11 MiB/sec sha3-224: 62.86 MiB/sec sha3-256: 60.22 MiB/sec sha3-384: 45.69 MiB/sec sha3-512: 28.58 MiB/sec
4k随机读写测试：
[root@CloudDemo-9904 fio-master]# fio -filename=/root/bench/test.fio -direct=1 -rw=rw -bs=4k -size 1G -numjobs=8 -runtime=30 -group_reporting -name=file
file: (g=0): rw=rw, bs=4K-4K/4K-4K/4K-4K, ioengine=sync, iodepth=1
...
file: (g=0): rw=rw, bs=4K-4K/4K-4K/4K-4K, ioengine=sync, iodepth=1
fio-2.0.13
Starting 8 processes
file: Laying out IO file(s) (1 file(s) / 1024MB)
Jobs: 8 (f=8): [MMMMMMMM] [100.0% done] [1752K/1756K/0K /s] [438 /439 /0 iops] [eta 00m:00s]
file: (groupid=0, jobs=8): err= 0: pid=31890: Thu Jul 6 22:22:45 2017
read : io=70080KB, bw=2334.2KB/s, iops=583 , runt= 30013msec
clat (usec): min=38 , max=196410 , avg=6120.59, stdev=11528.88
lat (usec): min=38 , max=196410 , avg=6120.83, stdev=11528.85
clat percentiles (usec):
| 1.00th=[ 52], 5.00th=[ 61], 10.00th=[ 71], 20.00th=[ 88],
| 30.00th=[ 111], 40.00th=[ 153], 50.00th=[ 342], 60.00th=[ 3664],
| 70.00th=[ 6880], 80.00th=[10688], 90.00th=[17280], 95.00th=[24960],
| 99.00th=[52480], 99.50th=[70144], 99.90th=[122368], 99.95th=[138240],
| 99.99th=[164864]
bw (KB/s) : min= 14, max= 546, per=12.57%, avg=293.32, stdev=107.41
write: io=70240KB, bw=2340.4KB/s, iops=585 , runt= 30013msec
clat (usec): min=461 , max=183868 , avg=7551.08, stdev=12332.16
lat (usec): min=461 , max=183868 , avg=7551.52, stdev=12332.17
clat percentiles (usec):
| 1.00th=[ 532], 5.00th=[ 596], 10.00th=[ 644], 20.00th=[ 756],
| 30.00th=[ 900], 40.00th=[ 1256], 50.00th=[ 2896], 60.00th=[ 5280],
| 70.00th=[ 8256], 80.00th=[11968], 90.00th=[19072], 95.00th=[27008],
| 99.00th=[58624], 99.50th=[79360], 99.90th=[134144], 99.95th=[150528],
| 99.99th=[177152]
bw (KB/s) : min= 7, max= 603, per=12.57%, avg=294.14, stdev=111.43
lat (usec) : 50=0.30%, 100=12.49%, 250=10.74%, 500=3.15%, 750=10.44%
lat (usec) : 1000=7.56%
lat (msec) : 2=6.52%, 4=6.76%, 10=18.87%, 20=14.79%, 50=7.19%
lat (msec) : 100=0.95%, 250=0.25%
cpu : usr=0.10%, sys=0.59%, ctx=59868, majf=1, minf=221
IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0%
submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
issued : total=r=17520/w=17560/d=0, short=r=0/w=0/d=0
Run status group 0 (all jobs):
READ: io=70080KB, aggrb=2334KB/s, minb=2334KB/s, maxb=2334KB/s, mint=30013msec, maxt=30013msec
WRITE: io=70240KB, aggrb=2340KB/s, minb=2340KB/s, maxb=2340KB/s, mint=30013msec, maxt=30013msec
Disk stats (read/write):
sda: ios=17444/17489, merge=0/7, ticks=3339/28319, in_queue=31638, util=96.63%
32k随机读写测试：
[root@CloudDemo-9904 fio-master]# fio -filename=/root/bench/test.fio -direct=1 -rw=rw -bs=32k -size 1G -numjobs=8 -runtime=30 -group_reporting -name=file
file: (g=0): rw=rw, bs=32K-32K/32K-32K/32K-32K, ioengine=sync, iodepth=1
...
file: (g=0): rw=rw, bs=32K-32K/32K-32K/32K-32K, ioengine=sync, iodepth=1
fio-2.0.13
Starting 8 processes
Jobs: 8 (f=8): [MMMMMMMM] [100.0% done] [14304K/16416K/0K /s] [447 /513 /0 iops] [eta 00m:00s]
file: (groupid=0, jobs=8): err= 0: pid=31900: Thu Jul 6 22:23:34 2017
read : io=469376KB, bw=15641KB/s, iops=488 , runt= 30010msec
clat (usec): min=47 , max=179832 , avg=7268.80, stdev=11956.80
lat (usec): min=47 , max=179832 , avg=7269.06, stdev=11956.78
clat percentiles (usec):
| 1.00th=[ 65], 5.00th=[ 76], 10.00th=[ 85], 20.00th=[ 104],
| 30.00th=[ 137], 40.00th=[ 215], 50.00th=[ 628], 60.00th=[ 5280],
| 70.00th=[ 9024], 80.00th=[13376], 90.00th=[20608], 95.00th=[28288],
| 99.00th=[56064], 99.50th=[68096], 99.90th=[110080], 99.95th=[122368],
| 99.99th=[175104]
bw (KB/s) : min= 617, max= 3351, per=12.53%, avg=1960.11, stdev=557.52
write: io=480576KB, bw=16014KB/s, iops=500 , runt= 30010msec
clat (usec): min=660 , max=185935 , avg=8872.92, stdev=12785.60
lat (usec): min=663 , max=185938 , avg=8876.01, stdev=12785.44
clat percentiles (usec):
| 1.00th=[ 716], 5.00th=[ 788], 10.00th=[ 852], 20.00th=[ 996],
| 30.00th=[ 1208], 40.00th=[ 1800], 50.00th=[ 3536], 60.00th=[ 6880],
| 70.00th=[10432], 80.00th=[14784], 90.00th=[22144], 95.00th=[30848],
| 99.00th=[61696], 99.50th=[79360], 99.90th=[116224], 99.95th=[142336],
| 99.99th=[175104]
bw (KB/s) : min= 377, max= 3812, per=12.53%, avg=2006.20, stdev=649.03
lat (usec) : 50=0.02%, 100=8.82%, 250=11.86%, 500=2.71%, 750=3.07%
lat (usec) : 1000=9.48%
lat (msec) : 2=11.86%, 4=6.68%, 10=16.12%, 20=18.10%, 50=9.82%
lat (msec) : 100=1.30%, 250=0.18%
cpu : usr=0.12%, sys=0.55%, ctx=51348, majf=0, minf=238
IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0%
submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
issued : total=r=14668/w=15018/d=0, short=r=0/w=0/d=0
Run status group 0 (all jobs):
READ: io=469376KB, aggrb=15640KB/s, minb=15640KB/s, maxb=15640KB/s, mint=30010msec, maxt=30010msec
WRITE: io=480576KB, aggrb=16013KB/s, minb=16013KB/s, maxb=16013KB/s, mint=30010msec, maxt=30010msec
Disk stats (read/write):
sda: ios=14625/14984, merge=0/10, ticks=3782/28755, in_queue=32515, util=97.24%
1m随机读写测试：
[root@CloudDemo-9904 fio-master]# fio -filename=/root/bench/test.fio -direct=1 -rw=rw -bs=1m -size 1G -numjobs=8 -runtime=30 -group_reporting -name=file
file: (g=0): rw=rw, bs=1M-1M/1M-1M/1M-1M, ioengine=sync, iodepth=1
...
file: (g=0): rw=rw, bs=1M-1M/1M-1M/1M-1M, ioengine=sync, iodepth=1
fio-2.0.13
Starting 8 processes
Jobs: 8 (f=8): [MMMMMMMM] [100.0% done] [92252K/89177K/0K /s] [90 /87 /0 iops] [eta 00m:00s]
file: (groupid=0, jobs=8): err= 0: pid=31910: Thu Jul 6 22:24:11 2017
read : io=1912.0MB, bw=65081KB/s, iops=63 , runt= 30084msec
clat (usec): min=517 , max=519191 , avg=55901.56, stdev=62410.36
lat (usec): min=517 , max=519191 , avg=55902.21, stdev=62410.29
clat percentiles (usec):
| 1.00th=[ 572], 5.00th=[ 724], 10.00th=[ 1096], 20.00th=[ 1384],
| 30.00th=[ 2064], 40.00th=[10432], 50.00th=[42752], 60.00th=[62720],
| 70.00th=[81408], 80.00th=[102912], 90.00th=[138240], 95.00th=[166912],
| 99.00th=[248832], 99.50th=[292864], 99.90th=[440320], 99.95th=[518144],
| 99.99th=[518144]
bw (KB/s) : min= 1098, max=19387, per=12.84%, avg=8357.64, stdev=3822.84
write: io=1889.0MB, bw=64298KB/s, iops=62 , runt= 30084msec
clat (msec): min=5 , max=514 , avg=70.48, stdev=68.08
lat (msec): min=5 , max=514 , avg=70.68, stdev=68.06
clat percentiles (msec):
| 1.00th=[ 7], 5.00th=[ 8], 10.00th=[ 10], 20.00th=[ 13],
| 30.00th=[ 17], 40.00th=[ 26], 50.00th=[ 55], 60.00th=[ 77],
| 70.00th=[ 95], 80.00th=[ 120], 90.00th=[ 159], 95.00th=[ 198],
| 99.00th=[ 277], 99.50th=[ 379], 99.90th=[ 502], 99.95th=[ 515],
| 99.99th=[ 515]
bw (KB/s) : min= 1490, max=22061, per=12.85%, avg=8263.35, stdev=4358.54
lat (usec) : 750=2.60%, 1000=1.16%
lat (msec) : 2=11.00%, 4=2.87%, 10=8.95%, 20=12.76%, 50=11.65%
lat (msec) : 100=24.41%, 250=23.26%, 500=1.26%, 750=0.08%
cpu : usr=0.18%, sys=0.24%, ctx=7276, majf=0, minf=238
IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0%
submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
issued : total=r=1912/w=1889/d=0, short=r=0/w=0/d=0
Run status group 0 (all jobs):
READ: io=1912.0MB, aggrb=65080KB/s, minb=65080KB/s, maxb=65080KB/s, mint=30084msec, maxt=30084msec
WRITE: io=1889.0MB, aggrb=64297KB/s, minb=64297KB/s, maxb=64297KB/s, mint=30084msec, maxt=30084msec
Disk stats (read/write):
sda: ios=3804/3773, merge=0/3, ticks=9523/55479, in_queue=65008, util=98.68%
因为没有用上SSD，所以数据都很难看。不过我注意到他们的云主机有存储优化型主机，不知道磁盘会不会用上SSD。不过月付25美元，就算不是存储优化型，也不应该那么慢，至少也要用上SSD缓存才对。
0x04.3 ramsmp
RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes INTEGER & WRITING 1 Kb block: 14903.08 MB/s INTEGER & WRITING 2 Kb block: 15857.77 MB/s INTEGER & WRITING 4 Kb block: 15198.78 MB/s INTEGER & WRITING 8 Kb block: 16392.72 MB/s INTEGER & WRITING 16 Kb block: 15491.18 MB/s INTEGER & WRITING 32 Kb block: 15944.29 MB/s INTEGER & WRITING 64 Kb block: 14994.42 MB/s INTEGER & WRITING 128 Kb block: 15040.56 MB/s INTEGER & WRITING 256 Kb block: 14568.11 MB/s INTEGER & WRITING 512 Kb block: 13336.18 MB/s INTEGER & WRITING 1024 Kb block: 13306.06 MB/s INTEGER & WRITING 2048 Kb block: 12831.18 MB/s INTEGER & WRITING 4096 Kb block: 10846.68 MB/s INTEGER & WRITING 8192 Kb block: 8145.52 MB/s INTEGER & WRITING 16384 Kb block: 6726.50 MB/s INTEGER & WRITING 32768 Kb block: 6735.43 MB/s RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes INTEGER & READING 1 Kb block: 32941.98 MB/s INTEGER & READING 2 Kb block: 33584.97 MB/s INTEGER & READING 4 Kb block: 32618.16 MB/s INTEGER & READING 8 Kb block: 34872.57 MB/s INTEGER & READING 16 Kb block: 35798.44 MB/s INTEGER & READING 32 Kb block: 33816.48 MB/s INTEGER & READING 64 Kb block: 23455.91 MB/s INTEGER & READING 128 Kb block: 21855.56 MB/s INTEGER & READING 256 Kb block: 19286.62 MB/s INTEGER & READING 512 Kb block: 18520.86 MB/s INTEGER & READING 1024 Kb block: 16327.69 MB/s INTEGER & READING 2048 Kb block: 15418.05 MB/s INTEGER & READING 4096 Kb block: 14227.11 MB/s INTEGER & READING 8192 Kb block: 10258.68 MB/s INTEGER & READING 16384 Kb block: 7061.02 MB/s INTEGER & READING 32768 Kb block: 7393.16 MB/s RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes INTEGER Copy: 7572.41 MB/s INTEGER Scale: 8264.90 MB/s INTEGER Add: 7623.15 MB/s INTEGER Triad: 7631.20 MB/s --- INTEGER AVERAGE: 7772.92 MB/s RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes FL-POINT & WRITING 1 Kb block: 15567.11 MB/s FL-POINT & WRITING 2 Kb block: 15013.28 MB/s FL-POINT & WRITING 4 Kb block: 16030.97 MB/s FL-POINT & WRITING 8 Kb block: 14101.41 MB/s FL-POINT & WRITING 16 Kb block: 16789.16 MB/s FL-POINT & WRITING 32 Kb block: 15058.03 MB/s FL-POINT & WRITING 64 Kb block: 14119.14 MB/s FL-POINT & WRITING 128 Kb block: 14613.80 MB/s FL-POINT & WRITING 256 Kb block: 13280.61 MB/s FL-POINT & WRITING 512 Kb block: 13262.02 MB/s FL-POINT & WRITING 1024 Kb block: 12066.01 MB/s FL-POINT & WRITING 2048 Kb block: 12033.12 MB/s FL-POINT & WRITING 4096 Kb block: 10724.89 MB/s FL-POINT & WRITING 8192 Kb block: 8281.63 MB/s FL-POINT & WRITING 16384 Kb block: 6699.45 MB/s FL-POINT & WRITING 32768 Kb block: 6712.22 MB/s RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes FL-POINT & READING 1 Kb block: 17483.03 MB/s FL-POINT & READING 2 Kb block: 17787.60 MB/s FL-POINT & READING 4 Kb block: 18010.26 MB/s FL-POINT & READING 8 Kb block: 16200.30 MB/s FL-POINT & READING 16 Kb block: 17309.19 MB/s FL-POINT & READING 32 Kb block: 17628.80 MB/s FL-POINT & READING 64 Kb block: 17630.74 MB/s FL-POINT & READING 128 Kb block: 16993.51 MB/s FL-POINT & READING 256 Kb block: 16777.61 MB/s FL-POINT & READING 512 Kb block: 16486.17 MB/s FL-POINT & READING 1024 Kb block: 15413.94 MB/s FL-POINT & READING 2048 Kb block: 14523.41 MB/s FL-POINT & READING 4096 Kb block: 13901.60 MB/s FL-POINT & READING 8192 Kb block: 9919.73 MB/s FL-POINT & READING 16384 Kb block: 7912.85 MB/s FL-POINT & READING 32768 Kb block: 8006.13 MB/s RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes FL-POINT Copy: 6008.81 MB/s FL-POINT Scale: 5790.92 MB/s FL-POINT Add: 7638.70 MB/s FL-POINT Triad: 7516.54 MB/s --- FL-POINT AVERAGE: 6738.74 MB/s RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes MMX & WRITING 1 Kb block: 15351.07 MB/s MMX & WRITING 2 Kb block: 16004.67 MB/s MMX & WRITING 4 Kb block: 15340.16 MB/s MMX & WRITING 8 Kb block: 16516.57 MB/s MMX & WRITING 16 Kb block: 15902.94 MB/s MMX & WRITING 32 Kb block: 17049.17 MB/s MMX & WRITING 64 Kb block: 13849.87 MB/s MMX & WRITING 128 Kb block: 13411.93 MB/s MMX & WRITING 256 Kb block: 14089.16 MB/s MMX & WRITING 512 Kb block: 12814.78 MB/s MMX & WRITING 1024 Kb block: 13063.69 MB/s MMX & WRITING 2048 Kb block: 11598.55 MB/s MMX & WRITING 4096 Kb block: 9697.15 MB/s MMX & WRITING 8192 Kb block: 8385.40 MB/s MMX & WRITING 16384 Kb block: 6956.12 MB/s MMX & WRITING 32768 Kb block: 6559.50 MB/s RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes MMX & READING 1 Kb block: 35238.12 MB/s MMX & READING 2 Kb block: 35139.27 MB/s MMX & READING 4 Kb block: 24705.76 MB/s MMX & READING 8 Kb block: 35152.43 MB/s MMX & READING 16 Kb block: 35001.10 MB/s MMX & READING 32 Kb block: 36748.03 MB/s MMX & READING 64 Kb block: 26677.98 MB/s MMX & READING 128 Kb block: 36406.98 MB/s MMX & READING 256 Kb block: 35573.43 MB/s MMX & READING 512 Kb block: 27720.85 MB/s MMX & READING 1024 Kb block: 33828.56 MB/s MMX & READING 2048 Kb block: 26928.43 MB/s MMX & READING 4096 Kb block: 23773.71 MB/s MMX & READING 8192 Kb block: 16970.17 MB/s MMX & READING 16384 Kb block: 9508.35 MB/s MMX & READING 32768 Kb block: 8781.46 MB/s RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes MMX Copy: 8478.51 MB/s MMX Scale: 8368.43 MB/s MMX Add: 8145.54 MB/s MMX Triad: 7707.17 MB/s --- MMX AVERAGE: 8174.91 MB/s RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes SSE & WRITING 1 Kb block: 35880.76 MB/s SSE & WRITING 2 Kb block: 36251.94 MB/s SSE & WRITING 4 Kb block: 31029.44 MB/s SSE & WRITING 8 Kb block: 31529.75 MB/s SSE & WRITING 16 Kb block: 34600.94 MB/s SSE & WRITING 32 Kb block: 26557.68 MB/s SSE & WRITING 64 Kb block: 23351.32 MB/s SSE & WRITING 128 Kb block: 20115.13 MB/s SSE & WRITING 256 Kb block: 18895.07 MB/s SSE & WRITING 512 Kb block: 16326.42 MB/s SSE & WRITING 1024 Kb block: 14128.01 MB/s SSE & WRITING 2048 Kb block: 14759.78 MB/s SSE & WRITING 4096 Kb block: 13733.61 MB/s SSE & WRITING 8192 Kb block: 9231.10 MB/s SSE & WRITING 16384 Kb block: 6447.67 MB/s SSE & WRITING 32768 Kb block: 6843.68 MB/s RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes SSE & READING 1 Kb block: 73402.87 MB/s SSE & READING 2 Kb block: 73857.13 MB/s SSE & READING 4 Kb block: 72622.65 MB/s SSE & READING 8 Kb block: 42815.23 MB/s SSE & READING 16 Kb block: 74375.75 MB/s SSE & READING 32 Kb block: 72760.29 MB/s SSE & READING 64 Kb block: 73316.42 MB/s SSE & READING 128 Kb block: 73642.79 MB/s SSE & READING 256 Kb block: 40926.28 MB/s SSE & READING 512 Kb block: 69057.31 MB/s SSE & READING 1024 Kb block: 68843.12 MB/s SSE & READING 2048 Kb block: 66532.68 MB/s SSE & READING 4096 Kb block: 25473.49 MB/s SSE & READING 8192 Kb block: 18866.51 MB/s SSE & READING 16384 Kb block: 10125.65 MB/s SSE & READING 32768 Kb block: 9367.97 MB/s RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes SSE Copy: 8174.90 MB/s SSE Scale: 8496.06 MB/s SSE Add: 8056.44 MB/s SSE Triad: 8268.90 MB/s --- SSE AVERAGE: 8249.08 MB/s RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes MMX (nt) & WRITING 1 Kb block: 3972.49 MB/s MMX (nt) & WRITING 2 Kb block: 3790.87 MB/s MMX (nt) & WRITING 4 Kb block: 3963.87 MB/s MMX (nt) & WRITING 8 Kb block: 3801.55 MB/s MMX (nt) & WRITING 16 Kb block: 3930.67 MB/s MMX (nt) & WRITING 32 Kb block: 3907.39 MB/s MMX (nt) & WRITING 64 Kb block: 3916.14 MB/s MMX (nt) & WRITING 128 Kb block: 3967.85 MB/s MMX (nt) & WRITING 256 Kb block: 4184.10 MB/s MMX (nt) & WRITING 512 Kb block: 3987.48 MB/s MMX (nt) & WRITING 1024 Kb block: 3921.43 MB/s MMX (nt) & WRITING 2048 Kb block: 3973.96 MB/s MMX (nt) & WRITING 4096 Kb block: 3958.39 MB/s MMX (nt) & WRITING 8192 Kb block: 3931.79 MB/s MMX (nt) & WRITING 16384 Kb block: 3967.83 MB/s MMX (nt) & WRITING 32768 Kb block: 4117.83 MB/s RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes MMX (nt) & READING 1 Kb block: 34351.32 MB/s MMX (nt) & READING 2 Kb block: 24417.58 MB/s MMX (nt) & READING 4 Kb block: 32997.76 MB/s MMX (nt) & READING 8 Kb block: 33440.43 MB/s MMX (nt) & READING 16 Kb block: 27142.45 MB/s MMX (nt) & READING 32 Kb block: 32774.66 MB/s MMX (nt) & READING 64 Kb block: 24786.37 MB/s MMX (nt) & READING 128 Kb block: 32314.58 MB/s MMX (nt) & READING 256 Kb block: 32553.63 MB/s MMX (nt) & READING 512 Kb block: 29948.48 MB/s MMX (nt) & READING 1024 Kb block: 21634.46 MB/s MMX (nt) & READING 2048 Kb block: 29756.57 MB/s MMX (nt) & READING 4096 Kb block: 23668.23 MB/s MMX (nt) & READING 8192 Kb block: 20165.75 MB/s MMX (nt) & READING 16384 Kb block: 11012.75 MB/s MMX (nt) & READING 32768 Kb block: 9829.44 MB/s RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes MMX (nt) Copy: 5200.93 MB/s [NTA prefetch] MMX (nt) Scale: 5008.79 MB/s [NTA prefetch] MMX (nt) Add: 6348.31 MB/s [T0 prefetch] MMX (nt) Triad: 6150.64 MB/s [T0 prefetch] --- MMX (nt) AVERAGE: 5677.17 MB/s RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes SSE & WRITING (nt) 1 Kb block: 4140.29 MB/s SSE & WRITING (nt) 2 Kb block: 4141.11 MB/s SSE & WRITING (nt) 4 Kb block: 4204.65 MB/s SSE & WRITING (nt) 8 Kb block: 3960.09 MB/s SSE & WRITING (nt) 16 Kb block: 4264.69 MB/s SSE & WRITING (nt) 32 Kb block: 4202.97 MB/s SSE & WRITING (nt) 64 Kb block: 4092.03 MB/s SSE & WRITING (nt) 128 Kb block: 4148.24 MB/s SSE & WRITING (nt) 256 Kb block: 4300.12 MB/s SSE & WRITING (nt) 512 Kb block: 3993.07 MB/s SSE & WRITING (nt) 1024 Kb block: 4296.50 MB/s SSE & WRITING (nt) 2048 Kb block: 4191.57 MB/s SSE & WRITING (nt) 4096 Kb block: 4166.28 MB/s SSE & WRITING (nt) 8192 Kb block: 4288.35 MB/s SSE & WRITING (nt) 16384 Kb block: 4231.25 MB/s SSE & WRITING (nt) 32768 Kb block: 4069.79 MB/s RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes SSE & READING (nt) 1 Kb block: 62027.01 MB/s SSE & READING (nt) 2 Kb block: 61262.70 MB/s SSE & READING (nt) 4 Kb block: 45364.42 MB/s SSE & READING (nt) 8 Kb block: 58529.69 MB/s SSE & READING (nt) 16 Kb block: 57149.93 MB/s SSE & READING (nt) 32 Kb block: 56154.19 MB/s SSE & READING (nt) 64 Kb block: 42086.02 MB/s SSE & READING (nt) 128 Kb block: 55651.86 MB/s SSE & READING (nt) 256 Kb block: 55544.54 MB/s SSE & READING (nt) 512 Kb block: 50461.35 MB/s SSE & READING (nt) 1024 Kb block: 40723.04 MB/s SSE & READING (nt) 2048 Kb block: 52827.33 MB/s SSE & READING (nt) 4096 Kb block: 32641.06 MB/s SSE & READING (nt) 8192 Kb block: 22003.96 MB/s SSE & READING (nt) 16384 Kb block: 10680.48 MB/s SSE & READING (nt) 32768 Kb block: 9876.43 MB/s RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes SSE (nt) Copy: 6240.65 MB/s [NTA prefetch] SSE (nt) Scale: 6220.15 MB/s [NTA prefetch] SSE (nt) Add: 7099.28 MB/s [T0 prefetch] SSE (nt) Triad: 7557.46 MB/s [T0 prefetch] --- SSE (nt) AVERAGE: 6779.39 MB/s
内存部分还算不错，可以接受。
0x04.4 Unix Benchmark
# # # # # # # ##### ###### # # #### # #
# # ## # # # # # # # ## # # # # #
# # # # # # ## ##### ##### # # # # ######
# # # # # # ## # # # # # # # # #
# # # ## # # # # # # # ## # # # #
#### # # # # # ##### ###### # # #### # #
Version 5.1.3 Based on the Byte Magazine Unix Benchmark
Multi-CPU version Version 5 revisions by Ian Smith,
Sunnyvale, CA, USA
January 13, 2011 johantheghost at yahoo period com
1 x Dhrystone 2 using register variables 1 2 3 4 5 6 7 8 9 10
1 x Double-Precision Whetstone 1 2 3 4 5 6 7 8 9 10
1 x Execl Throughput 1 2 3
1 x File Copy 1024 bufsize 2000 maxblocks 1 2 3
1 x File Copy 256 bufsize 500 maxblocks 1 2 3
1 x File Copy 4096 bufsize 8000 maxblocks 1 2 3
1 x Pipe Throughput 1 2 3 4 5 6 7 8 9 10
1 x Pipe-based Context Switching 1 2 3 4 5 6 7 8 9 10
1 x Process Creation 1 2 3
1 x System Call Overhead 1 2 3 4 5 6 7 8 9 10
1 x Shell Scripts (1 concurrent) 1 2 3
1 x Shell Scripts (8 concurrent) 1 2 3
========================================================================
BYTE UNIX Benchmarks (Version 5.1.3)
System: CloudDemo-9904: GNU/Linux
OS: GNU/Linux -- 2.6.32-696.el6.x86_64 -- #1 SMP Tue Mar 21 19:29:05 UTC 2017
Machine: x86_64 (x86_64)
Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
CPU 0: Intel(R) Xeon(R) CPU E5-2620 v2 @ 2.10GHz (4200.0 bogomips)
x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
22:40:52 up 2 days, 12:05, 1 user, load average: 0.10, 0.56, 0.85; runlevel 2017-07-04
------------------------------------------------------------------------
Benchmark Run: Thu Jul 06 2017 22:40:52 - 23:09:01
1 CPU in system; running 1 parallel copy of tests
Dhrystone 2 using register variables 25019171.0 lps (10.0 s, 7 samples)
Double-Precision Whetstone 3141.2 MWIPS (9.4 s, 7 samples)
Execl Throughput 3085.1 lps (29.8 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks 527400.5 KBps (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks 167457.9 KBps (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks 1315569.2 KBps (30.0 s, 2 samples)
Pipe Throughput 1436849.1 lps (10.0 s, 7 samples)
Pipe-based Context Switching 262893.1 lps (10.0 s, 7 samples)
Process Creation 9398.6 lps (30.1 s, 2 samples)
Shell Scripts (1 concurrent) 3379.3 lpm (60.0 s, 2 samples)
Shell Scripts (8 concurrent) 445.7 lpm (60.0 s, 2 samples)
System Call Overhead 1743205.3 lps (10.0 s, 7 samples)
System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 25019171.0 2143.9
Double-Precision Whetstone 55.0 3141.2 571.1
Execl Throughput 43.0 3085.1 717.5
File Copy 1024 bufsize 2000 maxblocks 3960.0 527400.5 1331.8
File Copy 256 bufsize 500 maxblocks 1655.0 167457.9 1011.8
File Copy 4096 bufsize 8000 maxblocks 5800.0 1315569.2 2268.2
Pipe Throughput 12440.0 1436849.1 1155.0
Pipe-based Context Switching 4000.0 262893.1 657.2
Process Creation 126.0 9398.6 745.9
Shell Scripts (1 concurrent) 42.4 3379.3 797.0
Shell Scripts (8 concurrent) 6.0 445.7 742.8
System Call Overhead 15000.0 1743205.3 1162.1
========
System Benchmarks Index Score 1003.8
评分刚好破千，但我还是觉得不值25美元。
0x05 结语
网络和磁盘评价一家服务商是否优质的重要因素，krypt给出这个价格，对于这个性能，月付25美元确实太贵了。如果月付25美元可以加上CN2线路，我想大家都会抢着买。
以下是krypt独立服务器的评测，买的是最低配置，所以性能并不是很好。不过我主要想测试网络，美国机房走的时候BPG线路，虽然直连国内，可是电信依旧存在QOS，速度非常缓慢：
0x06 相关视频
https://www.bilibili.com/video/av11826670/
