Web Server Performance Benchmarks: Post yours here for comparison
 

Just benchmarked two servers, a VPS I have some stuff on and a new server from Razor that I'm testing out. Hoping some people here will run this benchmark as well and post their results. I shut off apache before I ran the tests on my VPS to keep things somewhat fair, so it was under basically no load, and the new server is a fresh install.

I'll have results from a $20 linode VPS later tonight on a friend's box that I'll post here as well.

To install this you'll need gcc, so apt-get install gcc or yum install gcc or whatever

wget http://members.dslextreme.com/users/...1.0-wht.tar.gz
gunzip -dvc unixbench-4.1.0-wht.tar.gz | tar xvf -
cd unixbench-4.1.0-wht-2
make all
./Run

Tests took 8-10 minutes for me.

Linode 1536: (1.5 gb ram VPS) $60 a month

================================================== ============
BYTE UNIX Benchmarks (Version 4.1-wht.2)
System -- Linux localhost 3.0.18-linode43 #1 SMP Mon Jan 30 11:44:09 EST 2012 i686 GNU/Linux
/dev/xvda 57798908 52652840 4558868 93% /

Start Benchmark Run: Tue Feb 28 17:58:20 EST 2012
17:58:20 up 8 days, 10:17, 1 user, load average: 0.17, 0.22, 0.24

End Benchmark Run: Tue Feb 28 18:08:28 EST 2012
18:08:28 up 8 days, 10:27, 1 user, load average: 17.15, 6.69, 2.97


INDEX VALUES
TEST BASELINE RESULT INDEX

Dhrystone 2 using register variables 376783.7 18743196.1 497.5
Double-Precision Whetstone 83.1 1192.2 143.5
Execl Throughput 188.3 3567.8 189.5
File Copy 1024 bufsize 2000 maxblocks 2672.0 43536.0 162.9
File Copy 256 bufsize 500 maxblocks 1077.0 9461.0 87.8
File Read 4096 bufsize 8000 maxblocks 15382.0 311346.0 202.4
Pipe Throughput 111814.6 164662.3 14.7
Pipe-based Context Switching 15448.6 75162.2 48.7
Process Creation 569.3 6722.4 118.1
Shell Scripts (8 concurrent) 44.8 1251.8 279.4
System Call Overhead 114433.5 1865715.8 163.0
=========
FINAL SCORE 128.0

Razor: Xeon E31230 / 8gb ram $99/mo

================================================== ============
BYTE UNIX Benchmarks (Version 4.1-wht.2)
System -- Linux 3.0.0-16-generic #28-Ubuntu SMP Fri Jan 27 17:44:39 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux
/dev/sda5 480387816 1086968 454898520 1% /

Start Benchmark Run: Tue Feb 28 17:56:30 EST 2012
17:56:30 up 28 min, 1 user, load average: 0.26, 0.13, 0.08

End Benchmark Run: Tue Feb 28 18:06:41 EST 2012
18:06:40 up 38 min, 1 user, load average: 9.56, 4.19, 1.93


INDEX VALUES
TEST BASELINE RESULT INDEX

Dhrystone 2 using register variables 376783.7 53699353.9 1425.2
Double-Precision Whetstone 83.1 2568.5 309.1
Execl Throughput 188.3 10128.4 537.9
File Copy 1024 bufsize 2000 maxblocks 2672.0 995765.0 3726.7
File Copy 256 bufsize 500 maxblocks 1077.0 313081.0 2907.0
File Read 4096 bufsize 8000 maxblocks 15382.0 4174985.0 2714.2
Pipe-based Context Switching 15448.6 349372.0 226.2
Pipe Throughput 111814.6 9609502.5 859.4
Process Creation 569.3 27283.7 479.2
Shell Scripts (8 concurrent) 44.8 8133.0 1815.4
System Call Overhead 114433.5 16971419.2 1483.1
=========
FINAL SCORE 1052.0

Not going to do all of that...

But with my old host I was sitting around 3.x to 7.x server load which meant my surfers were fucked...

And my new host which is probably in my sig took that down to .04 so they must be doing something right.

Aw c'mon man humour me, it's like 5 commands you can copy and paste lol.

I might just end up at your host. I like the specs of this server, and I know razor has nice new hardware in a good building, but their support is like another world compared to my linode VPS. I don't think I ever waited more than 10 minutes for a support ticket there, and here it looks like there isn't anyone in the building at night. Phone just rings, tickets don't get answered.

I'm fine lowering my expectations there, it's a different business model entirely, I understand, but I don't know if I can live with it. Also simple shit like DNS was a couple of text boxes and clicking save with linode, and here it's (I think!) manually creating zone files and an ugly as fuck interface for adding DNS entries.. See: http://support.razorservers.com/inde...=kb.page&id=11

Not really sure what to do there, I guess just add the www entry and do all the cname mx record stuff in my zone files. Shrug.

All I can say is our main server loads faster then 92% of the internet on average. Kick Ass 24/7 baby :thumbsup:thumbsup

ChicagoVPS 2 GB RAM / 50 GB HDD / 2 TB bandwith for 7$ / month + 5$ for Cpanel (VPS):

another node on ChicagoVPS 2 GB RAM / 50 GB HDD / 2 TB bandwith for 7$ / month + 5$ for Cpanel (VPS):

Both servers are currently handling about 400-500 reqs/min. :thumbsup

By the way for 7$ use coupon code '2048' and after activation use addon for cPanel and use 'cPanel" for 10$ recurring discount on cPanel HERE ;)

Kimsufi 24G from http://www.kimsufi.co.uk/
Intel i7 with 24 gb ram for $85 a month

Dhrystone 2 using register variables 376783.7 32434155.2 860.8
Double-Precision Whetstone 83.1 2146.5 258.3
Execl Throughput 188.3 14564.9 773.5
File Copy 1024 bufsize 2000 maxblocks 2672.0 628460.0 2352.0
File Copy 256 bufsize 500 maxblocks 1077.0 195927.0 1819.2
File Read 4096 bufsize 8000 maxblocks 15382.0 3594869.0 2337.1
Pipe-based Context Switching 15448.6 917662.7 594.0
Pipe Throughput 111814.6 8597073.8 768.9
Process Creation 569.3 45637.0 801.6
Shell Scripts (8 concurrent) 44.8 4936.0 1101.8
System Call Overhead 114433.5 15344848.6 1340.9
=========
FINAL SCORE 993.5

INDEX VALUES
TEST BASELINE RESULT INDEX
Some machine:

Dhrystone 2 using register variables 376783.7 21582360.4 572.8
Double-Precision Whetstone 83.1 1813.4 218.2
Execl Throughput 188.3 9614.4 510.6
File Copy 1024 bufsize 2000 maxblocks 2672.0 154296.0 577.5
File Copy 256 bufsize 500 maxblocks 1077.0 37285.0 346.2
File Read 4096 bufsize 8000 maxblocks 15382.0 1496704.0 973.0
Pipe-based Context Switching 15448.6 651168.8 421.5
Pipe Throughput 111814.6 3203261.1 286.5
Process Creation 569.3 34496.4 605.9
Shell Scripts (8 concurrent) 44.8 4453.8 994.2
System Call Overhead 114433.5 3078839.2 269.1
=========
FINAL SCORE 469.3

What machine? :)

2 x Quadcore Xeons 2.5ghz with 16gb ram

Hm, so my VPSes are not that bad with scores similar to yours :)

not necessarily .. the biggest bottleneck with VPSes is the I/O which this doesnt test

But on these grounds - what this benchmark tests, I got a similar result than others with dedi :)

I think what's giving me the good scores is the solid processor that comes with this package the E31230, ranks highly here: http://www.cpubenchmark.net/high_end_cpus.html