网络

YardVPS Tree-1 性能测试

留下评论

VPS 都用了这么久了才想起来跑下 UnixBench ，有点失策哈。

YardVPS Tree-1 售价为 7.95 刀/月（差不多 ￥55/月 的样子），宣称的配置为：

  • 512MB RAM + 1GB Swap
  • 1 CPU Core
  • 15GB Disk Space （新用户现在有 20GB 了，我们老用户真可怜 T.T ）
  • 100Mbps Network （对此我持怀疑态度）
  • 99.99% Uptime （这个绝对是扯淡）

这些都是硬指标，那看不见的那部分性能如何？服务器上装的是 Ubuntu 10.04.3 LTS ，内核则是 Linux dword1511.info 2.6.32-340-ec2 #40-Ubuntu SMP Wed Nov 16 14:36:38 UTC 2011 x86_64 GNU/Linux ，下面直接奉上结果：

========================================================================
   BYTE UNIX Benchmarks (Version 5.1.3)

   System: dword1511.info: GNU/Linux
   OS: GNU/Linux -- 2.6.32-340-ec2 -- #40-Ubuntu SMP Wed Nov 16 14:36:38 UTC 2011
   Machine: x86_64 (unknown)
   Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
   CPU 0: Intel(R) Xeon(R) CPU E5620 @ 2.40GHz (4818.2 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   13:45:15 up 23 days,  7:00,  1 user,  load average: 0.22, 0.12, 0.09; runlevel 2

------------------------------------------------------------------------
Benchmark Run: 一  2月 06 2012 13:45:15 - 14:13:50
1 CPU in system; running 1 parallel copy of tests

Dhrystone 2 using register variables       17499972.4 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     2574.8 MWIPS (9.9 s, 7 samples)
Execl Throughput                               1096.1 lps   (29.8 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        176000.3 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks           48668.5 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks        550077.4 KBps  (30.0 s, 2 samples)
Pipe Throughput                              304126.8 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                  69843.2 lps   (10.0 s, 7 samples)
Process Creation                               2428.1 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   1954.3 lpm   (60.1 s, 2 samples)
Shell Scripts (8 concurrent)                    239.1 lpm   (60.2 s, 2 samples)
System Call Overhead                         348200.0 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   17499972.4   1499.6
Double-Precision Whetstone                       55.0       2574.8    468.1
Execl Throughput                                 43.0       1096.1    254.9
File Copy 1024 bufsize 2000 maxblocks          3960.0     176000.3    444.4
File Copy 256 bufsize 500 maxblocks            1655.0      48668.5    294.1
File Copy 4096 bufsize 8000 maxblocks          5800.0     550077.4    948.4
Pipe Throughput                               12440.0     304126.8    244.5
Pipe-based Context Switching                   4000.0      69843.2    174.6
Process Creation                                126.0       2428.1    192.7
Shell Scripts (1 concurrent)                     42.4       1954.3    460.9
Shell Scripts (8 concurrent)                      6.0        239.1    398.5
System Call Overhead                          15000.0     348200.0    232.1
                                                                   ========
System Benchmarks Index Score                                         375.2

这个分数略土啊，一般 Xen 的 VPS 跑到 700+ （不同版本的 UnixBench 综合指数的 baseline 不一样，我用的是 Byte UnixBench 。）都是很轻松的事情。不过今天晚上 VPS 状态不好也是真的， I/O 感觉非常的 laggy ，也许隔壁有人在跑大杀器也不一定。

高分主要集中在整数运算跟大文件读写这块。

对比下我的笔电（仅供参考）：

========================================================================
   BYTE UNIX Benchmarks (Version 5.1.3)

   System: dword-as4741g: GNU/Linux
   OS: GNU/Linux -- 2.6.38-13-generic -- #53-Ubuntu SMP Mon Nov 28 19:33:45 UTC 2011
   Machine: x86_64 (x86_64)
   Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
   CPU 0: Intel(R) Core(TM) i5 CPU M 430 @ 2.27GHz (4522.3 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET, Intel virtualization
   CPU 1: Intel(R) Core(TM) i5 CPU M 430 @ 2.27GHz (4522.0 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET, Intel virtualization
   CPU 2: Intel(R) Core(TM) i5 CPU M 430 @ 2.27GHz (4522.0 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET, Intel virtualization
   CPU 3: Intel(R) Core(TM) i5 CPU M 430 @ 2.27GHz (4522.0 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET, Intel virtualization
   22:38:38 up 11 days,  6:34,  5 users,  load average: 1.72, 1.98, 2.06; runlevel 2

------------------------------------------------------------------------
Benchmark Run: 一  2月 06 2012 22:38:38 - 23:07:43
4 CPUs in system; running 1 parallel copy of tests

Dhrystone 2 using register variables       24884988.3 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     3109.2 MWIPS (9.9 s, 7 samples)
Execl Throughput                               1692.7 lps   (29.7 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        725935.4 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks          251511.4 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks       1416855.0 KBps  (30.0 s, 2 samples)
Pipe Throughput                             1704106.3 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                  83979.2 lps   (10.0 s, 7 samples)
Process Creation                               5821.3 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   2596.6 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                   1081.5 lpm   (60.1 s, 2 samples)
System Call Overhead                        3156469.1 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   24884988.3   2132.4
Double-Precision Whetstone                       55.0       3109.2    565.3
Execl Throughput                                 43.0       1692.7    393.7
File Copy 1024 bufsize 2000 maxblocks          3960.0     725935.4   1833.2
File Copy 256 bufsize 500 maxblocks            1655.0     251511.4   1519.7
File Copy 4096 bufsize 8000 maxblocks          5800.0    1416855.0   2442.9
Pipe Throughput                               12440.0    1704106.3   1369.9
Pipe-based Context Switching                   4000.0      83979.2    209.9
Process Creation                                126.0       5821.3    462.0
Shell Scripts (1 concurrent)                     42.4       2596.6    612.4
Shell Scripts (8 concurrent)                      6.0       1081.5   1802.4
System Call Overhead                          15000.0    3156469.1   2104.3
                                                                   ========
System Benchmarks Index Score                                         999.7

------------------------------------------------------------------------
Benchmark Run: 一  2月 06 2012 23:07:43 - 23:37:01
4 CPUs in system; running 4 parallel copies of tests

Dhrystone 2 using register variables       53847546.1 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                    10486.0 MWIPS (9.9 s, 7 samples)
Execl Throughput                              10195.2 lps   (30.0 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        690758.9 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks          189838.2 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks       1752344.9 KBps  (30.0 s, 2 samples)
Pipe Throughput                             3976690.4 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                 625011.4 lps   (10.0 s, 7 samples)
Process Creation                              29849.2 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   9116.7 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                   1171.1 lpm   (60.1 s, 2 samples)
System Call Overhead                        7961857.7 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   53847546.1   4614.2
Double-Precision Whetstone                       55.0      10486.0   1906.5
Execl Throughput                                 43.0      10195.2   2371.0
File Copy 1024 bufsize 2000 maxblocks          3960.0     690758.9   1744.3
File Copy 256 bufsize 500 maxblocks            1655.0     189838.2   1147.1
File Copy 4096 bufsize 8000 maxblocks          5800.0    1752344.9   3021.3
Pipe Throughput                               12440.0    3976690.4   3196.7
Pipe-based Context Switching                   4000.0     625011.4   1562.5
Process Creation                                126.0      29849.2   2369.0
Shell Scripts (1 concurrent)                     42.4       9116.7   2150.2
Shell Scripts (8 concurrent)                      6.0       1171.1   1951.8
System Call Overhead                          15000.0    7961857.7   5307.9
                                                                   ========
System Benchmarks Index Score                                        2381.2

可以看到单核综合指数有 999.7 ，双核四线程则是 2381.2 ，比 YardVPS 的服务器快不少。

笔电的具体情况：

  • Acer Aspire 4741G 432G32Mn
  • 硬盘换成 ST9750420AS
  • 内存为创见 DDR3 1333 × 2
  • 电源适配器在线，房间温度约 10℃
  • Intel Core i5 430M + HM55
  • Ubuntu 11.04
  • gcc (Ubuntu/Linaro 4.5.2-8ubuntu4) 4.5.2
  • Linux dword-as4741g 2.6.38-13-generic #53-Ubuntu SMP Mon Nov 28 19:33:45 UTC 2011 x86_64 x86_64 x86_64 GNU/Linux

顺便再次吐槽下 YardVPS 的稳定性， VPS 的 Uptime 才到笔电的 2 倍有木有！

发表回复

您的电子邮箱地址不会被公开。 必填项已用*标注

此站点使用Akismet来减少垃圾评论。了解我们如何处理您的评论数据。