0x01 前言
我目前使用pachosting的云服务器，但在过去的半年中出现了3到4次服务中断的情况，其中有2次中断长达8小时。但是他们的网络状况非常好，无论是通过电信、联通还是移动访问，延迟都在20ms以下，因为他们的数据中心位于中国香港，同时使用光纤直连内地的方式提供服务。
我目前用的这个服务器快到期了，正打算续费，在查找付费按钮的时候发现原先的云服务器产品已经停用，取而代之的是新品牌layerstack。以下是layerstack的相关地址：
0x02 产品
layerstack目前包含有两种服务：
- 基于Openstack的云服务器
- Acronis异地备份
在这里我仅测试云服务器的部分，最便宜的一档年付报价2256港币，比品牌升级之前的2376港币略为便宜，相关配置信息与价格如下：
需要注意的是，以上价格为年付/12的价格，如果需要月付，价格会略微上涨。
在购买之前你可以申请试用，申请过程中可能会接到客服的确认电话，请注意填写正确的个人信息。通过审核的话，登入信息会通过邮件发送到邮箱里。
pachosting销售客服的服务时间为周一到周五，早上9点到下午7点；周六早上9点到下午6点；周日休息。因为我是老客户，所以申请6小时候就收到了使用机器的信息。
0x03 界面
控制台界面没有其他花俏的功能：
目前支持的系统的版本都比较新，目前还没找到自行上传系统镜像的地方，还不知道后期会不会添加这功能或添加更多不同版本的系统：
另外还多了一个之前没有的功能，在网络选项卡中，可以在网络层面配置防火墙了，如果不想在系统中配置防火墙，这功能还是挺好的。默认情况下是放行所有流量的：
0x04 网络
0x04.1 ping
因为是光纤直连中国内地，所以延迟非常低，以下是ping测试结果：
--- 146.196.52.24 ping statistics --- 100 packets transmitted, 99 received, 1% packet loss, time 99110ms rtt min/avg/max/mdev = 16.633/26.311/333.944/33.132 ms
0x04.2 wget
以下是下载测试：
[root@web temp]# wget http://146.196.52.24/1gb.test --2017-04-24 17:08:42-- http://146.196.52.24/1gb.test Connecting to 146.196.52.24:80... connected. HTTP request sent, awaiting response... 200 OK Length: 1073741824 (1.0G) [application/octet-stream] Saving to: ‘1gb.test’ 1073741824 1.36MB/s 11m 10s
下载测试中完全跑满了我的带宽，12M电缆，而且速度维持在1.35m/s以上。
0x04.3 traceroute
以下是路由追踪测试：
[root@web temp]# traceroute 146.196.52.24 traceroute to 146.196.52.24 (146.196.52.24), 30 hops max, 60 byte packets 1 pfSense.t.com (10.1.1.1) 0.219 ms 0.221 ms 0.302 ms 2 113.89.192.1 (113.89.192.1) 10.676 ms 10.707 ms 12.413 ms 3 113.106.44.105 (113.106.44.105) 15.651 ms 16.091 ms 18.083 ms 4 183.56.65.10 (183.56.65.10) 25.354 ms 183.56.65.2 (183.56.65.2) 25.700 ms 25.849 ms 5 59.43.80.17 (59.43.80.17) 27.352 ms 29.571 ms 202.97.26.50 (202.97.26.50) 36.364 ms 6 59.43.19.54 (59.43.19.54) 32.925 ms 59.43.80.17 (59.43.80.17) 32.767 ms 35.422 ms 7 59.43.19.54 (59.43.19.54) 47.459 ms 12.657 ms * 8 59.43.246.134 (59.43.246.134) 15.538 ms 59.43.183.106 (59.43.183.106) 22.677 ms 22.679 ms 9 59.43.247.230 (59.43.247.230) 26.388 ms 59.43.183.106 (59.43.183.106) 27.581 ms 27.680 ms 10 59.43.247.230 (59.43.247.230) 30.217 ms 32.641 ms 210.48.137.50 (210.48.137.50) 36.008 ms 11 * * 210.48.137.50 (210.48.137.50) 37.504 ms 12 * 146.196.52.24 (146.196.52.24) 41.835 ms 44.451 ms
从上面的数据可以看出，第10条还是国内CN2线路，第11跳已经到香港，第12跳就到了服务器，非常棒的线路。
以下使用webkaka进行测试：
0x04.4 speedtest
ping测试我想应该是不太可信的，另外各个测试节点的带宽也不尽相同，所以会有误差。
[root@enginx speedtest-cli]# python speedtest.py --server 5081 Retrieving speedtest.net configuration... Testing from (146.196.52.24)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by ShenZhen Telecom (Shenzhen) [31.69 km]: 608.182 ms Testing download speed................................................................................ Download: 200.18 Mbit/s Testing upload speed.................................................................................................... Upload: 90.57 Mbit/s [root@enginx speedtest-cli]# python speedtest.py --server 10201 Retrieving speedtest.net configuration... Testing from (146.196.52.24)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by China Network Unicom Shenzhen Branch (Shenzhen) [31.69 km]: 1052.106 ms Testing download speed................................................................................ Download: 85.49 Mbit/s Testing upload speed.................................................................................................... Upload: 37.07 Mbit/s [root@enginx speedtest-cli]# python speedtest.py --server 5506 Retrieving speedtest.net configuration... Testing from (146.196.52.24)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by china unicom (Xiamen) [468.88 km]: 54.961 ms Testing download speed................................................................................ Download: 116.15 Mbit/s Testing upload speed.................................................................................................... Upload: 39.43 Mbit/s [root@enginx speedtest-cli]# python speedtest.py --server 10192 Retrieving speedtest.net configuration... Testing from (146.196.52.24)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by NanningTelecom (Nanning) [604.08 km]: 52.563 ms Testing download speed................................................................................ Download: 200.21 Mbit/s Testing upload speed.................................................................................................... Upload: 46.69 Mbit/s [root@enginx speedtest-cli]# python speedtest.py --server 10305 Retrieving speedtest.net configuration... Testing from (146.196.52.24)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by GX-Telecom (Nanning) [604.08 km]: 49.713 ms Testing download speed................................................................................ Download: 182.55 Mbit/s Testing upload speed.................................................................................................... Upload: 54.32 Mbit/s [root@enginx speedtest-cli]# python speedtest.py --server 4870 Retrieving speedtest.net configuration... Testing from (146.196.52.24)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by Changsha, Hunan Unicom (Changsha) [671.99 km]: 41.517 ms Testing download speed................................................................................ Download: 79.17 Mbit/s Testing upload speed.................................................................................................... Upload: 32.13 Mbit/s [root@enginx speedtest-cli]# python speedtest.py --server 6132 Retrieving speedtest.net configuration... Testing from (146.196.52.24)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by ChinaTelecom.Hunan (Changsha) [671.99 km]: 737.99 ms Testing download speed................................................................................ Download: 204.88 Mbit/s Testing upload speed.................................................................................................... Upload: 52.95 Mbit/s [root@enginx speedtest-cli]# python speedtest.py --server 4863 Retrieving speedtest.net configuration... Testing from (146.196.52.24)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by Xi'an branch,Chinaunicom (Xi'an) [1431.66 km]: 598.493 ms Testing download speed................................................................................ Download: 48.31 Mbit/s Testing upload speed.................................................................................................... Upload: 26.01 Mbit/s [root@enginx speedtest-cli]# python speedtest.py --server 5292 Retrieving speedtest.net configuration... Testing from (146.196.52.24)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by China Mobile Group Shaanxi Company Limited (Xi'an) [1431.66 km]: 115.362 ms Testing download speed................................................................................ Download: 65.82 Mbit/s Testing upload speed.................................................................................................... Upload: 24.51 Mbit/s [root@enginx speedtest-cli]# python speedtest.py --server 4884 Retrieving speedtest.net configuration... Testing from (146.196.52.24)... Retrieving speedtest.net server list... Selecting best server based on ping... Hosted by China Unicom FuJian (Fuzhou) [672.83 km]: 401.732 ms Testing download speed................................................................................ Download: 154.85 Mbit/s Testing upload speed.................................................................................................... Upload: 34.76 Mbit/s
以下是测试数据的简单图表，从表中可以看出，下行速度高达200m/s，而上行速度可以达到20~100m/s
0x05 性能
我这台云服务器的配置如下：
- 架构：KVM
- CPU：2.4GHz x 4核心
- 内存：4GB
- 硬盘：150GB SSD
- 流量：无限，带宽未标明
- IP地址：1个 ipv4
0x05.1 基于dd的硬盘IO测试
5次测试结果如下：
[root@enginx temp]# 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, 1.02588 s, 1.0 GB/s [root@enginx temp]# 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, 1.03157 s, 1.0 GB/s [root@enginx temp]# 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, 1.01131 s, 1.1 GB/s [root@enginx temp]# 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, 0.922518 s, 1.2 GB/s [root@enginx temp]# 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, 1.02098 s, 1.1 GB/s
0x05.2 基于fio的IO测试
以下是加解密运算测试：
[root@enginx ~]# fio -crctest md5: 426.93 MiB/sec crc64: 278.29 MiB/sec crc32: 244.67 MiB/sec crc32c: 278.61 MiB/sec crc16: 281.69 MiB/sec crc7: 282.98 MiB/sec sha1: 381.32 MiB/sec sha256: 162.56 MiB/sec sha512: 269.30 MiB/sec xxhash: 4462.73 MiB/sec murmur3: 1804.84 MiB/sec jhash: 1776.67 MiB/sec fnv: 4538.60 MiB/sec sha3-224: 91.26 MiB/sec sha3-256: 86.19 MiB/sec sha3-384: 66.18 MiB/sec sha3-512: 45.23 MiB/sec
以下是混合顺序读写IO测试（4k）：
[root@enginx ~]# 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=(R) 4096B-4096B, (W) 4096B-4096B, (T) 4096B-4096B, ioengine=psync, iodepth=1
...
fio-2.19-13-gd7658
Starting 8 processes
Jobs: 8 (f=8): [M(8)][100.0%][r=82.2MiB/s,w=82.3MiB/s][r=21.3k,w=21.5k IOPS][eta 00m:00s]
file: (groupid=0, jobs=8): err= 0: pid=21240: Mon Apr 24 10:41:53 2017
read: IOPS=20.4k, BW=79.7MiB/s (83.6MB/s)(2391MiB/30001msec)
clat (usec): min=53, max=9707, avg=203.37, stdev=92.71
lat (usec): min=53, max=9707, avg=203.61, stdev=92.71
clat percentiles (usec):
| 1.00th=[ 77], 5.00th=[ 91], 10.00th=[ 100], 20.00th=[ 117],
| 30.00th=[ 153], 40.00th=[ 187], 50.00th=[ 209], 60.00th=[ 227],
| 70.00th=[ 247], 80.00th=[ 266], 90.00th=[ 294], 95.00th=[ 322],
| 99.00th=[ 374], 99.50th=[ 406], 99.90th=[ 1048], 99.95th=[ 1320],
| 99.99th=[ 2384]
write: IOPS=20.5k, BW=79.1MiB/s (83.8MB/s)(2397MiB/30001msec)
clat (usec): min=51, max=6232, avg=185.08, stdev=87.43
lat (usec): min=51, max=6232, avg=185.44, stdev=87.42
clat percentiles (usec):
| 1.00th=[ 67], 5.00th=[ 78], 10.00th=[ 87], 20.00th=[ 102],
| 30.00th=[ 131], 40.00th=[ 165], 50.00th=[ 189], 60.00th=[ 209],
| 70.00th=[ 227], 80.00th=[ 249], 90.00th=[ 278], 95.00th=[ 302],
| 99.00th=[ 358], 99.50th=[ 386], 99.90th=[ 860], 99.95th=[ 1272],
| 99.99th=[ 2352]
lat (usec) : 100=14.20%, 250=62.15%, 500=23.45%, 750=0.08%, 1000=0.02%
lat (msec) : 2=0.08%, 4=0.02%, 10=0.01%
cpu : usr=1.08%, sys=3.98%, ctx=1241389, majf=0, minf=259
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 rwt: total=611996,613588,0, short=0,0,0, dropped=0,0,0
latency : target=0, window=0, percentile=100.00%, depth=1
Run status group 0 (all jobs):
READ: bw=79.7MiB/s (83.6MB/s), 79.7MiB/s-79.7MiB/s (83.6MB/s-83.6MB/s), io=2391MiB (2507MB), run=30001-30001msec
WRITE: bw=79.1MiB/s (83.8MB/s), 79.1MiB/s-79.1MiB/s (83.8MB/s-83.8MB/s), io=2397MiB (2513MB), run=30001-30001msec
Disk stats (read/write):
vda: ios=609923/611441, merge=0/0, ticks=119602/108275, in_queue=228296, util=100.00%
以下是混合顺序读写IO测试（32k）：
[root@enginx ~]# 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=(R) 32.0KiB-32.0KiB, (W) 32.0KiB-32.0KiB, (T) 32.0KiB-32.0KiB, ioengine=psync, iodepth=1
...
fio-2.19-13-gd7658
Starting 8 processes
Jobs: 8 (f=8): [M(8)][100.0%][r=542MiB/s,w=540MiB/s][r=17.4k,w=17.3k IOPS][eta 00m:00s]
file: (groupid=0, jobs=8): err= 0: pid=21250: Mon Apr 24 10:43:19 2017
read: IOPS=17.3k, BW=538MiB/s (564MB/s)(4095MiB/7618msec)
clat (usec): min=65, max=7147, avg=237.86, stdev=122.05
lat (usec): min=65, max=7147, avg=238.11, stdev=122.06
clat percentiles (usec):
| 1.00th=[ 84], 5.00th=[ 96], 10.00th=[ 103], 20.00th=[ 117],
| 30.00th=[ 149], 40.00th=[ 213], 50.00th=[ 247], 60.00th=[ 274],
| 70.00th=[ 298], 80.00th=[ 326], 90.00th=[ 362], 95.00th=[ 398],
| 99.00th=[ 478], 99.50th=[ 524], 99.90th=[ 1352], 99.95th=[ 1512],
| 99.99th=[ 2512]
write: IOPS=17.3k, BW=538MiB/s (564MB/s)(4097MiB/7618msec)
clat (usec): min=65, max=5907, avg=220.53, stdev=118.00
lat (usec): min=67, max=5908, avg=221.74, stdev=118.01
clat percentiles (usec):
| 1.00th=[ 79], 5.00th=[ 88], 10.00th=[ 94], 20.00th=[ 105],
| 30.00th=[ 125], 40.00th=[ 191], 50.00th=[ 229], 60.00th=[ 255],
| 70.00th=[ 282], 80.00th=[ 310], 90.00th=[ 346], 95.00th=[ 378],
| 99.00th=[ 458], 99.50th=[ 502], 99.90th=[ 1304], 99.95th=[ 1464],
| 99.99th=[ 2448]
lat (usec) : 100=11.27%, 250=43.32%, 500=44.81%, 750=0.39%, 1000=0.03%
lat (msec) : 2=0.17%, 4=0.02%, 10=0.01%
cpu : usr=1.10%, sys=4.03%, ctx=264418, majf=0, minf=259
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 rwt: total=131049,131095,0, short=0,0,0, dropped=0,0,0
latency : target=0, window=0, percentile=100.00%, depth=1
Run status group 0 (all jobs):
READ: bw=538MiB/s (564MB/s), 538MiB/s-538MiB/s (564MB/s-564MB/s), io=4095MiB (4294MB), run=7618-7618msec
WRITE: bw=538MiB/s (564MB/s), 538MiB/s-538MiB/s (564MB/s-564MB/s), io=4097MiB (4296MB), run=7618-7618msec
Disk stats (read/write):
vda: ios=127505/127610, merge=0/0, ticks=29322/27113, in_queue=56602, util=99.08%
以下是混合顺序读写IO测试（1m）：
[root@enginx ~]# 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=(R) 1024KiB-1024KiB, (W) 1024KiB-1024KiB, (T) 1024KiB-1024KiB, ioengine=psync, iodepth=1
...
fio-2.19-13-gd7658
Starting 8 processes
Jobs: 8 (f=8)
file: (groupid=0, jobs=8): err= 0: pid=21260: Mon Apr 24 10:44:26 2017
read: IOPS=1921, BW=1921MiB/s (2015MB/s)(4127MiB/2148msec)
clat (usec): min=429, max=14218, avg=2077.53, stdev=1162.36
lat (usec): min=429, max=14218, avg=2077.84, stdev=1162.37
clat percentiles (usec):
| 1.00th=[ 470], 5.00th=[ 502], 10.00th=[ 524], 20.00th=[ 572],
| 30.00th=[ 956], 40.00th=[ 2128], 50.00th=[ 2352], 60.00th=[ 2512],
| 70.00th=[ 2672], 80.00th=[ 2896], 90.00th=[ 3248], 95.00th=[ 3696],
| 99.00th=[ 5024], 99.50th=[ 5408], 99.90th=[ 5984], 99.95th=[ 6176],
| 99.99th=[14272]
write: IOPS=1892, BW=1892MiB/s (1984MB/s)(4065MiB/2148msec)
clat (usec): min=399, max=6333, avg=2028.80, stdev=1068.27
lat (usec): min=424, max=6358, avg=2067.97, stdev=1068.65
clat percentiles (usec):
| 1.00th=[ 422], 5.00th=[ 474], 10.00th=[ 564], 20.00th=[ 628],
| 30.00th=[ 1496], 40.00th=[ 2128], 50.00th=[ 2288], 60.00th=[ 2448],
| 70.00th=[ 2608], 80.00th=[ 2800], 90.00th=[ 3120], 95.00th=[ 3504],
| 99.00th=[ 4896], 99.50th=[ 5216], 99.90th=[ 5664], 99.95th=[ 5792],
| 99.99th=[ 6304]
lat (usec) : 500=5.51%, 750=21.96%, 1000=2.15%
lat (msec) : 2=6.21%, 4=60.78%, 10=3.37%, 20=0.02%
cpu : usr=1.03%, sys=2.71%, ctx=8204, majf=0, minf=259
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 rwt: total=4127,4065,0, short=0,0,0, dropped=0,0,0
latency : target=0, window=0, percentile=100.00%, depth=1
Run status group 0 (all jobs):
READ: bw=1921MiB/s (2015MB/s), 1921MiB/s-1921MiB/s (2015MB/s-2015MB/s), io=4127MiB (4327MB), run=2148-2148msec
WRITE: bw=1892MiB/s (1984MB/s), 1892MiB/s-1892MiB/s (1984MB/s-1984MB/s), io=4065MiB (4262MB), run=2148-2148msec
Disk stats (read/write):
vda: ios=12378/12192, merge=0/0, ticks=25058/24261, in_queue=49296, util=95.63%
0x05.3 基于RAMSpeed/SMP的内存测试
[root@enginx ramsmp-3.5.0]# ./ramsmp -b 1 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: 37593.49 MB/s INTEGER & WRITING 2 Kb block: 37231.72 MB/s INTEGER & WRITING 4 Kb block: 37722.51 MB/s INTEGER & WRITING 8 Kb block: 37612.34 MB/s INTEGER & WRITING 16 Kb block: 36790.41 MB/s INTEGER & WRITING 32 Kb block: 37102.10 MB/s INTEGER & WRITING 64 Kb block: 32948.49 MB/s INTEGER & WRITING 128 Kb block: 33587.04 MB/s INTEGER & WRITING 256 Kb block: 31979.74 MB/s INTEGER & WRITING 512 Kb block: 30329.01 MB/s INTEGER & WRITING 1024 Kb block: 30106.48 MB/s INTEGER & WRITING 2048 Kb block: 29928.31 MB/s INTEGER & WRITING 4096 Kb block: 30126.17 MB/s INTEGER & WRITING 8192 Kb block: 29800.11 MB/s INTEGER & WRITING 16384 Kb block: 18747.05 MB/s INTEGER & WRITING 32768 Kb block: 14080.16 MB/s [root@enginx ramsmp-3.5.0]# ./ramsmp -b 2 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: 64911.63 MB/s INTEGER & READING 2 Kb block: 75662.08 MB/s INTEGER & READING 4 Kb block: 75772.03 MB/s INTEGER & READING 8 Kb block: 75497.49 MB/s INTEGER & READING 16 Kb block: 75204.35 MB/s INTEGER & READING 32 Kb block: 74417.64 MB/s INTEGER & READING 64 Kb block: 55847.61 MB/s INTEGER & READING 128 Kb block: 55684.24 MB/s INTEGER & READING 256 Kb block: 50795.40 MB/s INTEGER & READING 512 Kb block: 47898.28 MB/s INTEGER & READING 1024 Kb block: 47828.94 MB/s INTEGER & READING 2048 Kb block: 47816.56 MB/s INTEGER & READING 4096 Kb block: 48317.57 MB/s INTEGER & READING 8192 Kb block: 47345.27 MB/s INTEGER & READING 16384 Kb block: 27762.39 MB/s INTEGER & READING 32768 Kb block: 18093.99 MB/s [root@enginx ramsmp-3.5.0]# ./ramsmp -b 3 RAMspeed/SMP (Linux) v3.5.0 by Rhett M. Hollander and Paul V. Bolotoff, 2002-09 8Gb per pass mode, 2 processes INTEGER Copy: 17779.05 MB/s INTEGER Scale: 17679.38 MB/s INTEGER Add: 17114.63 MB/s INTEGER Triad: 17852.39 MB/s --- INTEGER AVERAGE: 17606.36 MB/s [root@enginx ramsmp-3.5.0]# ./ramsmp -b 4 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: 37349.08 MB/s FL-POINT & WRITING 2 Kb block: 37874.99 MB/s FL-POINT & WRITING 4 Kb block: 37850.63 MB/s FL-POINT & WRITING 8 Kb block: 38006.12 MB/s FL-POINT & WRITING 16 Kb block: 37896.80 MB/s FL-POINT & WRITING 32 Kb block: 37226.19 MB/s FL-POINT & WRITING 64 Kb block: 33600.01 MB/s FL-POINT & WRITING 128 Kb block: 33706.90 MB/s FL-POINT & WRITING 256 Kb block: 31972.01 MB/s FL-POINT & WRITING 512 Kb block: 30306.68 MB/s FL-POINT & WRITING 1024 Kb block: 30089.66 MB/s FL-POINT & WRITING 2048 Kb block: 30185.66 MB/s FL-POINT & WRITING 4096 Kb block: 30045.77 MB/s FL-POINT & WRITING 8192 Kb block: 29692.72 MB/s FL-POINT & WRITING 16384 Kb block: 16028.13 MB/s FL-POINT & WRITING 32768 Kb block: 14182.84 MB/s [root@enginx ramsmp-3.5.0]# ./ramsmp -b 5 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: 73564.75 MB/s FL-POINT & READING 2 Kb block: 74172.97 MB/s FL-POINT & READING 4 Kb block: 67811.48 MB/s FL-POINT & READING 8 Kb block: 73828.08 MB/s FL-POINT & READING 16 Kb block: 74004.27 MB/s FL-POINT & READING 32 Kb block: 74329.09 MB/s FL-POINT & READING 64 Kb block: 59269.75 MB/s FL-POINT & READING 128 Kb block: 57165.71 MB/s FL-POINT & READING 256 Kb block: 52066.52 MB/s FL-POINT & READING 512 Kb block: 49957.31 MB/s FL-POINT & READING 1024 Kb block: 50127.93 MB/s FL-POINT & READING 2048 Kb block: 50272.71 MB/s FL-POINT & READING 4096 Kb block: 50095.41 MB/s FL-POINT & READING 8192 Kb block: 48651.03 MB/s FL-POINT & READING 16384 Kb block: 49460.18 MB/s FL-POINT & READING 32768 Kb block: 19289.06 MB/s [root@enginx ramsmp-3.5.0]# ./ramsmp -b 6 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: 13896.60 MB/s FL-POINT Scale: 14237.57 MB/s FL-POINT Add: 16602.74 MB/s FL-POINT Triad: 17521.91 MB/s --- FL-POINT AVERAGE: 15564.71 MB/s
0x05.4 基于UnixBench的系统基准测试
[root@enginx ~]# cat codex/UnixBench/results/enginx-2017-04-24-01
BYTE UNIX Benchmarks (Version 5.1.3)
System: enginx: GNU/Linux
OS: GNU/Linux -- 3.10.0-514.6.2.el7.x86_64 -- #1 SMP Thu Feb 23 03:04:39 UTC 2017
Machine: x86_64 (x86_64)
Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
CPU 0: QEMU Virtual CPU version 2.5+ (4800.0 bogomips)
x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
CPU 1: QEMU Virtual CPU version 2.5+ (4800.0 bogomips)
x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
CPU 2: QEMU Virtual CPU version 2.5+ (4800.0 bogomips)
x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
CPU 3: QEMU Virtual CPU version 2.5+ (4800.0 bogomips)
x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
10:52:12 up 3:13, 1 user, load average: 0.04, 0.28, 0.34; runlevel 3
------------------------------------------------------------------------
Benchmark Run: Mon Apr 24 2017 10:52:12 - 11:20:15
4 CPUs in system; running 1 parallel copy of tests
Dhrystone 2 using register variables 28030781.2 lps (10.0 s, 7 samples)
Double-Precision Whetstone 4320.1 MWIPS (9.9 s, 7 samples)
Execl Throughput 3230.9 lps (30.0 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks 885325.2 KBps (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks 236824.2 KBps (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks 2536125.7 KBps (30.0 s, 2 samples)
Pipe Throughput 1300957.3 lps (10.0 s, 7 samples)
Pipe-based Context Switching 293564.5 lps (10.0 s, 7 samples)
Process Creation 10946.9 lps (30.0 s, 2 samples)
Shell Scripts (1 concurrent) 6984.1 lpm (60.0 s, 2 samples)
Shell Scripts (8 concurrent) 2500.6 lpm (60.0 s, 2 samples)
System Call Overhead 2131009.4 lps (10.0 s, 7 samples)
System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 28030781.2 2402.0
Double-Precision Whetstone 55.0 4320.1 785.5
Execl Throughput 43.0 3230.9 751.4
File Copy 1024 bufsize 2000 maxblocks 3960.0 885325.2 2235.7
File Copy 256 bufsize 500 maxblocks 1655.0 236824.2 1431.0
File Copy 4096 bufsize 8000 maxblocks 5800.0 2536125.7 4372.6
Pipe Throughput 12440.0 1300957.3 1045.8
Pipe-based Context Switching 4000.0 293564.5 733.9
Process Creation 126.0 10946.9 868.8
Shell Scripts (1 concurrent) 42.4 6984.1 1647.2
Shell Scripts (8 concurrent) 6.0 2500.6 4167.7
System Call Overhead 15000.0 2131009.4 1420.7
========
System Benchmarks Index Score 1499.2
------------------------------------------------------------------------
Benchmark Run: Mon Apr 24 2017 11:20:15 - 11:48:18
4 CPUs in system; running 4 parallel copies of tests
Dhrystone 2 using register variables 111437803.3 lps (10.0 s, 7 samples)
Double-Precision Whetstone 17282.1 MWIPS (9.8 s, 7 samples)
Execl Throughput 15175.1 lps (30.0 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks 1251302.2 KBps (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks 360872.5 KBps (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks 3930774.5 KBps (30.0 s, 2 samples)
Pipe Throughput 5208222.0 lps (10.0 s, 7 samples)
Pipe-based Context Switching 1178758.1 lps (10.0 s, 7 samples)
Process Creation 41833.7 lps (30.0 s, 2 samples)
Shell Scripts (1 concurrent) 21199.2 lpm (60.0 s, 2 samples)
Shell Scripts (8 concurrent) 2992.4 lpm (60.0 s, 2 samples)
System Call Overhead 6127322.5 lps (10.0 s, 7 samples)
System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 111437803.3 9549.1
Double-Precision Whetstone 55.0 17282.1 3142.2
Execl Throughput 43.0 15175.1 3529.1
File Copy 1024 bufsize 2000 maxblocks 3960.0 1251302.2 3159.9
File Copy 256 bufsize 500 maxblocks 1655.0 360872.5 2180.5
File Copy 4096 bufsize 8000 maxblocks 5800.0 3930774.5 6777.2
Pipe Throughput 12440.0 5208222.0 4186.7
Pipe-based Context Switching 4000.0 1178758.1 2946.9
Process Creation 126.0 41833.7 3320.1
Shell Scripts (1 concurrent) 42.4 21199.2 4999.8
Shell Scripts (8 concurrent) 6.0 2992.4 4987.4
System Call Overhead 15000.0 6127322.5 4084.9
========
System Benchmarks Index Score 4069.5
0x06 结语
经过一系列的测试，与上一代云产品相比较，layerstack这个产品的硬件性能和网络性能都较为优胜。但这一次测试所用服务器的配置较上一次测试的要高，所以不具备对比性。
以下是pachosting上一代云服务器产品的测试文章：
