ML115G1にnVraid5で仕込んだiscsiの設定ができたんで、WinXPpro32bitからつなげようと、下記をDownload.
microsoft iscsi software initiator 2.08
いやはや、繋がらないですね。ターゲット側の設定見直したりしましたが、おかしなところはなさそう。もうしらね、って思いお風呂入る前にWinXP再起動したら、あっさりつながった。インストールする時、再起動しなさいってメッセージなかったんだけどね~。ま、いいや。ハマった人、来てくれるかな。
で、早速ベンチマーク。
Sequential Read : 11.630 MB/s
Sequential Write : 8.577 MB/s
Random Read 512KB : 11.570 MB/s
Random Write 512KB : 6.136 MB/s
Random Read 4KB : 6.218 MB/s
Random Write 4KB : 0.185 MB/s
Test Size : 100 MB
Date : 2010/01/03 8:55:24
あれれ、10Mって。あ、HUBが10MのデスクトップスッチングHUBに繋がってました。基幹HUBより直結して再度測り直し。
Sequential Read : 107.878 MB/s
Sequential Write : 16.011 MB/s
Random Read 512KB : 103.359 MB/s
Random Write 512KB : 10.744 MB/s
Random Read 4KB : 20.320 MB/s
Random Write 4KB : 0.192 MB/s
Test Size : 100 MB
Date : 2010/01/03 9:02:30
ほほぉ、読むのは早いんだね。書き込みが遅いのはこんなもんか、設定の影響か。
ついでに、HDD TUNE2.55での結果も。
HD Tune: IET VIRTUAL-DISK Benchmark
Transfer Rate Minimum : 38.0 MB/sec
Transfer Rate Maximum : 78.0 MB/sec
Transfer Rate Average : 53.2 MB/sec
Access Time : 1.3 ms
Burst Rate : 64.3 MB/sec
CPU Usage : 8.4%
ちょっと遅い、昔のIDEディスクって感じですね。
テスト環境を下記に記しておきます。
▼クライアント側
WindowsXPpro32bit (ML115G5 AMD Phenom II X4 940 3GHz)
microsoft iscsi software initiator 2.08
NTFS 4Gbyte
network 1000Mbps (HUB FXG-0224TX)
▼サーバ側
CentOS5.4 x86-64 (ML115G1 AMD Opteron1216 2.4GHz)
iscsitarget-1.4.19
nvRaid5 (80G x 3 160G x1)あまりのディスクなのでバラバラですが
サーバ側はnvRaid MirrorとRaid5でそれほどかわりなかったので、Raid5にしました。以下、ベンチマーク。
==============================================================ML115 G1 Onbard Raid mirror
AMD Opteron1216 2.4GHz
CentOS5.4 x86_64
========================================================================
BYTE UNIX Benchmarks (Version 5.1.2)
System: ML115: GNU/Linux
OS: GNU/Linux — 2.6.18-164.el5 — #1 SMP Thu Sep 3 03:28:30 EDT 2009
Machine: x86_64 (x86_64)
Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
CPU 0: Dual-Core AMD Opteron(tm) Processor 1216 (2000.3 bogomips)
Hyper-Threading, x86-64, MMX, AMD MMX, Physical Address Ext, SYSENTER/SYSEXIT, AMD virtualization, SYSCALL/SYSRET
CPU 1: Dual-Core AMD Opteron(tm) Processor 1216 (2000.3 bogomips)
Hyper-Threading, x86-64, MMX, AMD MMX, Physical Address Ext, SYSENTER/SYSEXIT, AMD virtualization, SYSCALL/SYSRET
13:53:43 up 1:31, 2 users, load average: 0.00, 0.03, 0.35; runlevel 3
————————————————————————
Benchmark Run: Sat Jan 02 2010 13:53:43 - 14:21:23
2 CPUs in system; running 1 parallel copy of tests
Dhrystone 2 using register variables 7487416.2 lps (10.0 s, 7 samples)
Double-Precision Whetstone 1879.8 MWIPS (5.5 s, 7 samples)
Execl Throughput 1062.0 lps (29.7 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks 194820.7 KBps (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks 64547.5 KBps (30.1 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks 423057.0 KBps (30.0 s, 2 samples)
Pipe Throughput 478977.5 lps (10.0 s, 7 samples)
Pipe-based Context Switching 123755.0 lps (10.0 s, 7 samples)
Process Creation 8748.5 lps (30.0 s, 2 samples)
Shell Scripts (1 concurrent) 2027.2 lpm (60.0 s, 2 samples)
Shell Scripts (8 concurrent) 876.5 lpm (60.1 s, 2 samples)
System Call Overhead 857067.0 lps (10.0 s, 7 samples)
System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 7487416.2 641.6
Double-Precision Whetstone 55.0 1879.8 341.8
Execl Throughput 43.0 1062.0 247.0
File Copy 1024 bufsize 2000 maxblocks 3960.0 194820.7 492.0
File Copy 256 bufsize 500 maxblocks 1655.0 64547.5 390.0
File Copy 4096 bufsize 8000 maxblocks 5800.0 423057.0 729.4
Pipe Throughput 12440.0 478977.5 385.0
Pipe-based Context Switching 4000.0 123755.0 309.4
Process Creation 126.0 8748.5 694.3
Shell Scripts (1 concurrent) 42.4 2027.2 478.1
Shell Scripts (8 concurrent) 6.0 876.5 1460.9
System Call Overhead 15000.0 857067.0 571.4
========
System Benchmarks Index Score 501.0
————————————————————————
Benchmark Run: Sat Jan 02 2010 14:21:23 - 14:49:02
2 CPUs in system; running 2 parallel copies of tests
Dhrystone 2 using register variables 15136552.6 lps (10.0 s, 7 samples)
Double-Precision Whetstone 3744.1 MWIPS (4.9 s, 7 samples)
Execl Throughput 5154.2 lps (29.7 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks 117682.8 KBps (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks 23042.9 KBps (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks 430782.0 KBps (30.0 s, 2 samples)
Pipe Throughput 952388.3 lps (10.0 s, 7 samples)
Pipe-based Context Switching 309115.1 lps (10.0 s, 7 samples)
Process Creation 16014.6 lps (30.0 s, 2 samples)
Shell Scripts (1 concurrent) 6363.9 lpm (60.0 s, 2 samples)
Shell Scripts (8 concurrent) 918.0 lpm (60.0 s, 2 samples)
System Call Overhead 1413141.0 lps (10.0 s, 7 samples)
System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 15136552.6 1297.0
Double-Precision Whetstone 55.0 3744.1 680.8
Execl Throughput 43.0 5154.2 1198.6
File Copy 1024 bufsize 2000 maxblocks 3960.0 117682.8 297.2
File Copy 256 bufsize 500 maxblocks 1655.0 23042.9 139.2
File Copy 4096 bufsize 8000 maxblocks 5800.0 430782.0 742.7
Pipe Throughput 12440.0 952388.3 765.6
Pipe-based Context Switching 4000.0 309115.1 772.8
Process Creation 126.0 16014.6 1271.0
Shell Scripts (1 concurrent) 42.4 6363.9 1500.9
Shell Scripts (8 concurrent) 6.0 918.0 1529.9
System Call Overhead 15000.0 1413141.0 942.1
========
System Benchmarks Index Score 782.8
==============================================================ML115 G1 Onbard Raid 5
AMD Opteron1216 2.4GHz
CentOS5.4 x86_64
========================================================================
BYTE UNIX Benchmarks (Version 5.1.2)
System: ML115: GNU/Linux
OS: GNU/Linux — 2.6.18-164.el5 — #1 SMP Thu Sep 3 03:28:30 EDT 2009
Machine: x86_64 (x86_64)
Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
CPU 0: Dual-Core AMD Opteron(tm) Processor 1216 (2000.3 bogomips)
Hyper-Threading, x86-64, MMX, AMD MMX, Physical Address Ext, SYSENTER/SYSEXIT, AMD virtualization, SYSCALL/SYSRET
CPU 1: Dual-Core AMD Opteron(tm) Processor 1216 (2000.3 bogomips)
Hyper-Threading, x86-64, MMX, AMD MMX, Physical Address Ext, SYSENTER/SYSEXIT, AMD virtualization, SYSCALL/SYSRET
15:58:31 up 5 min, 1 user, load average: 0.11, 0.18, 0.09; runlevel 3
————————————————————————
Benchmark Run: Sat Jan 02 2010 15:58:31 - 16:26:08
2 CPUs in system; running 1 parallel copy of tests
Dhrystone 2 using register variables 7488099.2 lps (10.0 s, 7 samples)
Double-Precision Whetstone 1896.7 MWIPS (4.9 s, 7 samples)
Execl Throughput 1053.1 lps (29.7 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks 221188.4 KBps (30.1 s, 2 samples)
File Copy 256 bufsize 500 maxblocks 73972.3 KBps (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks 438177.3 KBps (30.0 s, 2 samples)
Pipe Throughput 495456.7 lps (10.0 s, 7 samples)
Pipe-based Context Switching 95076.0 lps (10.0 s, 7 samples)
Process Creation 8935.0 lps (30.0 s, 2 samples)
Shell Scripts (1 concurrent) 1929.5 lpm (60.0 s, 2 samples)
Shell Scripts (8 concurrent) 820.8 lpm (60.1 s, 2 samples)
System Call Overhead 904916.2 lps (10.0 s, 7 samples)
System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 7488099.2 641.7
Double-Precision Whetstone 55.0 1896.7 344.9
Execl Throughput 43.0 1053.1 244.9
File Copy 1024 bufsize 2000 maxblocks 3960.0 221188.4 558.6
File Copy 256 bufsize 500 maxblocks 1655.0 73972.3 447.0
File Copy 4096 bufsize 8000 maxblocks 5800.0 438177.3 755.5
Pipe Throughput 12440.0 495456.7 398.3
Pipe-based Context Switching 4000.0 95076.0 237.7
Process Creation 126.0 8935.0 709.1
Shell Scripts (1 concurrent) 42.4 1929.5 455.1
Shell Scripts (8 concurrent) 6.0 820.8 1368.0
System Call Overhead 15000.0 904916.2 603.3
========
System Benchmarks Index Score 502.3
————————————————————————
Benchmark Run: Sat Jan 02 2010 16:26:08 - 16:53:44
2 CPUs in system; running 2 parallel copies of tests
Dhrystone 2 using register variables 15025755.6 lps (10.0 s, 7 samples)
Double-Precision Whetstone 3976.3 MWIPS (4.6 s, 7 samples)
Execl Throughput 4989.9 lps (29.7 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks 87216.6 KBps (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks 23724.4 KBps (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks 433040.8 KBps (30.0 s, 2 samples)
Pipe Throughput 970015.0 lps (10.0 s, 7 samples)
Pipe-based Context Switching 309124.0 lps (10.0 s, 7 samples)
Process Creation 17086.1 lps (30.0 s, 2 samples)
Shell Scripts (1 concurrent) 5945.3 lpm (60.0 s, 2 samples)
Shell Scripts (8 concurrent) 853.6 lpm (60.1 s, 2 samples)
System Call Overhead 1464909.8 lps (10.0 s, 7 samples)
System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 15025755.6 1287.6
Double-Precision Whetstone 55.0 3976.3 723.0
Execl Throughput 43.0 4989.9 1160.4
File Copy 1024 bufsize 2000 maxblocks 3960.0 87216.6 220.2
File Copy 256 bufsize 500 maxblocks 1655.0 23724.4 143.3
File Copy 4096 bufsize 8000 maxblocks 5800.0 433040.8 746.6
Pipe Throughput 12440.0 970015.0 779.8
Pipe-based Context Switching 4000.0 309124.0 772.8
Process Creation 126.0 17086.1 1356.0
Shell Scripts (1 concurrent) 42.4 5945.3 1402.2
Shell Scripts (8 concurrent) 6.0 853.6 1422.6
System Call Overhead 15000.0 1464909.8 976.6
========
System Benchmarks Index Score 765.6
いやぁ、ML115って面白いマシンですね。