DL360G5 + LeftHand P4000G2 べんちまーく


某おもちゃ環境が遅すぎだろって思ったので、しみじみと計測してみた。

– 環境ていうか前提条件

Server:DL360G5
Storage:LeftHand P4000G2
SW:そこら辺で拾ったAlliedのバカHub(1Gbps)

ESX上で動いてるUbuntuにて計測、1CPUでメモリ1GBな環境。
DISKはLeftHand上に配置してます(´・ω・`)
空きポートが無くてNIC 1本づつしか繋げてない構成だけども、間違いなくネットワークがボトルネックになるケース。

– UNIX Benchmark等

とりあえず、全体的な性能を評価したかったので何も考えずUNIX Benchmarkしてみる。
VPSとかで各種結果とかがネットに転がってるので比較しやすい。

========================================================================
BYTE UNIX Benchmarks (Version 5.1.2)

System: penis-desktop: GNU/Linux
OS: GNU/Linux — 3.0.0-12-generic — #20-Ubuntu SMP Fri Oct 7 14:50:42 UTC 2011
Machine: i686 (i386)
Language: en_US.utf8 (charmap=”UTF-8″, collate=”UTF-8″)
CPU 0: Intel(R) Xeon(R) CPU 5148 @ 2.33GHz (4666.8 bogomips)
x86-64, MMX, Physical Address Ext
13:52:31 up 26 days, 18:45, 3 users, load average: 0.07, 0.03, 0.05; runlevel 2

————————————————————————
Benchmark Run: Thu Dec 22 2011 13:52:31 – 14:20:36
1 CPU in system; running 1 parallel copy of tests

Dhrystone 2 using register variables 11392862.4 lps (10.0 s, 7 samples)
Double-Precision Whetstone 2420.7 MWIPS (9.7 s, 7 samples)
Execl Throughput 1155.2 lps (30.0 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks 153775.8 KBps (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks 28540.5 KBps (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks 409697.2 KBps (30.0 s, 2 samples)
Pipe Throughput 137237.0 lps (10.0 s, 7 samples)
Pipe-based Context Switching 30117.2 lps (10.0 s, 7 samples)
Process Creation 1517.5 lps (30.0 s, 2 samples)
Shell Scripts (1 concurrent) 1404.0 lpm (60.0 s, 2 samples)
Shell Scripts (8 concurrent) 157.0 lpm (60.2 s, 2 samples)
System Call Overhead 220550.6 lps (10.0 s, 7 samples)

System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 11392862.4 976.3
Double-Precision Whetstone 55.0 2420.7 440.1
Execl Throughput 43.0 1155.2 268.7
File Copy 1024 bufsize 2000 maxblocks 3960.0 153775.8 388.3
File Copy 256 bufsize 500 maxblocks 1655.0 28540.5 172.5
File Copy 4096 bufsize 8000 maxblocks 5800.0 409697.2 706.4
Pipe Throughput 12440.0 137237.0 110.3
Pipe-based Context Switching 4000.0 30117.2 75.3
Process Creation 126.0 1517.5 120.4
Shell Scripts (1 concurrent) 42.4 1404.0 331.1
Shell Scripts (8 concurrent) 6.0 157.0 261.7
System Call Overhead 15000.0 220550.6 147.0
========
System Benchmarks Index Score 253.2

月額490円のServerManですら335.3ですよ。
自宅設置の5万円仮想箱でもSystem Benchmarks Index Score 1866.7なのに。

# hdparm -t /dev/sda
/dev/sda:
Timing buffered disk reads: 158 MB in 3.02 seconds = 52.30 MB/sec

所感ですけども、定説通りの感じですね。ネットワーク越しは最低でもTenGなNICが必須だと思うのでありますよ。
local >>> (越えられない壁) >>> iscsi-raw >> iscsi-image > nfs
あー遅い遅い( ゚∀゚)o彡°

Advertisements

About r00t_z3r0
最小で最大を

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: