View source for Benchmark for KVM Under Dell PowerEdge R710


Here is the benchmark result for a single-core virtual machine using Qemu-KVM hypervisor (in an OpenStack setup). The benchmark was performed under CentOS 7. The disk is served as a volume from cinder on the same machine serving data from a LVM volume on a Raid10.

<syntaxhighlight lang="text">
========================================================================
   BYTE UNIX Benchmarks (Version 5.1.3)

   System: centos-2: GNU/Linux
   OS: GNU/Linux -- 3.10.0-123.8.1.el7.x86_64 -- #1 SMP Mon Sep 22 19:06:58 UTC 2014
   Machine: x86_64 (x86_64)
   Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
   CPU 0: Westmere E56xx/L56xx/X56xx (Nehalem-C) (5320.0 bogomips)
          x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   20:57:57 up 39 min,  1 user,  load average: 0.00, 0.01, 0.05; runlevel 3

------------------------------------------------------------------------
Benchmark Run: Tue Nov 25 2014 20:57:57 - 21:26:03
1 CPU in system; running 1 parallel copy of tests

Dhrystone 2 using register variables       30583798.8 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     3929.5 MWIPS (10.0 s, 7 samples)
Execl Throughput                               3881.5 lps   (30.0 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks       1073559.0 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks          295659.6 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks       2292778.7 KBps  (30.0 s, 2 samples)
Pipe Throughput                             1643236.5 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                 312382.7 lps   (10.0 s, 7 samples)
Process Creation                              14438.5 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   3471.1 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                    775.5 lpm   (60.0 s, 2 samples)
System Call Overhead                        2217174.2 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   30583798.8   2620.7
Double-Precision Whetstone                       55.0       3929.5    714.4
Execl Throughput                                 43.0       3881.5    902.7
File Copy 1024 bufsize 2000 maxblocks          3960.0    1073559.0   2711.0
File Copy 256 bufsize 500 maxblocks            1655.0     295659.6   1786.5
File Copy 4096 bufsize 8000 maxblocks          5800.0    2292778.7   3953.1
Pipe Throughput                               12440.0    1643236.5   1320.9
Pipe-based Context Switching                   4000.0     312382.7    781.0
Process Creation                                126.0      14438.5   1145.9
Shell Scripts (1 concurrent)                     42.4       3471.1    818.7
Shell Scripts (8 concurrent)                      6.0        775.5   1292.5
System Call Overhead                          15000.0    2217174.2   1478.1
                                                                   ========
System Benchmarks Index Score                                        1406.3
</syntaxhighlight>

Here's the same test with the same distro but with 8 cores.

<syntaxhighlight lang="text">
========================================================================
   BYTE UNIX Benchmarks (Version 5.1.3)

   System: centos-3: GNU/Linux
   OS: GNU/Linux -- 3.10.0-123.8.1.el7.x86_64 -- #1 SMP Mon Sep 22 19:06:58 UTC 2014
   Machine: x86_64 (x86_64)
   Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
   CPU 0: Westmere E56xx/L56xx/X56xx (Nehalem-C) (5320.0 bogomips)
          x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 1: Westmere E56xx/L56xx/X56xx (Nehalem-C) (5320.0 bogomips)
          x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 2: Westmere E56xx/L56xx/X56xx (Nehalem-C) (5320.0 bogomips)
          x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 3: Westmere E56xx/L56xx/X56xx (Nehalem-C) (5320.0 bogomips)
          x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 4: Westmere E56xx/L56xx/X56xx (Nehalem-C) (5320.0 bogomips)
          x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 5: Westmere E56xx/L56xx/X56xx (Nehalem-C) (5320.0 bogomips)
          x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 6: Westmere E56xx/L56xx/X56xx (Nehalem-C) (5320.0 bogomips)
          x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 7: Westmere E56xx/L56xx/X56xx (Nehalem-C) (5320.0 bogomips)
          x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   21:39:51 up 7 min,  1 user,  load average: 0.78, 0.29, 0.12; runlevel 3

------------------------------------------------------------------------
Benchmark Run: Tue Nov 25 2014 21:39:51 - 22:08:03
8 CPUs in system; running 1 parallel copy of tests

Dhrystone 2 using register variables       30102305.5 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     3927.3 MWIPS (10.0 s, 7 samples)
Execl Throughput                               2377.9 lps   (30.0 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks       1093678.4 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks          301144.6 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks       2352469.5 KBps  (30.0 s, 2 samples)
Pipe Throughput                             1542204.9 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                 297146.7 lps   (10.0 s, 7 samples)
Process Creation                               9585.2 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   3684.2 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                   2562.3 lpm   (60.0 s, 2 samples)
System Call Overhead                        2079471.5 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   30102305.5   2579.5
Double-Precision Whetstone                       55.0       3927.3    714.0
Execl Throughput                                 43.0       2377.9    553.0
File Copy 1024 bufsize 2000 maxblocks          3960.0    1093678.4   2761.8
File Copy 256 bufsize 500 maxblocks            1655.0     301144.6   1819.6
File Copy 4096 bufsize 8000 maxblocks          5800.0    2352469.5   4056.0
Pipe Throughput                               12440.0    1542204.9   1239.7
Pipe-based Context Switching                   4000.0     297146.7    742.9
Process Creation                                126.0       9585.2    760.7
Shell Scripts (1 concurrent)                     42.4       3684.2    868.9
Shell Scripts (8 concurrent)                      6.0       2562.3   4270.4
System Call Overhead                          15000.0    2079471.5   1386.3
                                                                   ========
System Benchmarks Index Score                                        1432.7

------------------------------------------------------------------------
Benchmark Run: Tue Nov 25 2014 22:08:03 - 22:36:16
8 CPUs in system; running 8 parallel copies of tests

Dhrystone 2 using register variables      180872826.9 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                    27232.0 MWIPS (10.0 s, 7 samples)
Execl Throughput                              21915.9 lps   (30.0 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks       1628166.5 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks          476914.9 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks       3798316.0 KBps  (30.0 s, 2 samples)
Pipe Throughput                             9145378.7 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                1774389.1 lps   (10.0 s, 7 samples)
Process Creation                              56957.7 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                  33105.1 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                   4651.8 lpm   (60.0 s, 2 samples)
System Call Overhead                       10405920.4 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0  180872826.9  15499.0
Double-Precision Whetstone                       55.0      27232.0   4951.3
Execl Throughput                                 43.0      21915.9   5096.7
File Copy 1024 bufsize 2000 maxblocks          3960.0    1628166.5   4111.5
File Copy 256 bufsize 500 maxblocks            1655.0     476914.9   2881.7
File Copy 4096 bufsize 8000 maxblocks          5800.0    3798316.0   6548.8
Pipe Throughput                               12440.0    9145378.7   7351.6
Pipe-based Context Switching                   4000.0    1774389.1   4436.0
Process Creation                                126.0      56957.7   4520.4
Shell Scripts (1 concurrent)                     42.4      33105.1   7807.8
Shell Scripts (8 concurrent)                      6.0       4651.8   7753.0
System Call Overhead                          15000.0   10405920.4   6937.3
                                                                   ========
System Benchmarks Index Score                                        5926.8

</syntaxhighlight>

[[Category:Benchmarks]]