Litespeed vBulletin index.php benchmarks - Part 1

eva2000

Well-Known Member
#1
Just installed vB 3.8.6 PL1 test forum on test server to benchmark Litespeed v4.0.19. Only testing the front vB index.php page first.

Results are very much bottlenecked by MariaDB 5.2.4 MySQL server connection unlike standalone test.php or test.txt static file testing already done.

Summary
  • Litespeed again came out on top with least memory usage and serving up to 100,000 requests with 1,000 concurrency with no failed requests.
  • Nginx with php-fpm in TCP form came out in second place but had more failed requests as the concurrency and requests size increased. Update: Seems failed requests due to content length varying so not really failed requests see explanation here.
  • Nginx with php-fpm unix socket form for some reason couldn't even pass >5 concurrency level without php failing with 502 gateway error messages/upstream timeouts.
  • Apache kicked the bucket at 100,000 requests and cpu load was very high as was average memory usage.

Apachebench


:: Jan 28th, 2011: Correction to findings
I have to revise my findings a bit for Nginx. The failed requests seem to be from content length varying between apachebench requests and not due to failure requests related to connections/exceptions errors. Content length varies due to dynamic nature of php pages tested as as long as content length reported by apachebench as Document Length is nearly the same, it still means a successfully served request regardless. Full summary here


Raw numbers here


Server config
ran Unixbench 5.1.2 as per http://www.webhostingtalk.com/showthread.php?t=924581 to see how it compares.

  • VirtualBox CentOS 5.5 64bit Guest
  • Xeon W3540 @3408Mhz (Assigned 2 cores only) - linux reports @3375Mhz
  • 16MB allocated GPU memory
  • 1GB out of 6GB DDR3-1550Mhz
  • 20GB out of 640GB Samsung SATAII

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

   System: localhost.localdomain: GNU/Linux
   OS: GNU/Linux -- 2.6.18-194.32.1.el5 -- #1 SMP Wed Jan 5 17:52:25 EST 2011
   Machine: x86_64 (x86_64)
   Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
   CPU 0: Intel(R) Xeon(R) CPU W3540 @ 2.93GHz (6750.1 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 1: Intel(R) Xeon(R) CPU W3540 @ 2.93GHz (6733.8 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   21:45:27 up  1:11,  2 users,  load average: 0.00, 0.01, 0.00; runlevel 5

------------------------------------------------------------------------
Benchmark Run: Sun Jan 23 2011 21:45:27 - 22:09:32
2 CPUs in system; running 1 parallel copy of tests

Dhrystone 2 using register variables       19101079.5 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     3771.0 MWIPS (10.0 s, 7 samples)
Execl Throughput                               4394.2 lps   (30.0 s, 2 samples)
Pipe Throughput                             1457951.0 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                  29373.4 lps   (10.0 s, 7 samples)
Process Creation                              13113.2 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   7997.6 lpm   (60.0 s, 2 samples)
Shell Scripts (16 concurrent)                   901.5 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                   1756.9 lpm   (60.0 s, 2 samples)
System Call Overhead                        1211587.2 lps   (10.0 s, 7 samples)

System Benchmarks Partial Index              BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   19101079.5   1636.8
Double-Precision Whetstone                       55.0       3771.0    685.6
Execl Throughput                                 43.0       4394.2   1021.9
Pipe Throughput                               12440.0    1457951.0   1172.0
Pipe-based Context Switching                   4000.0      29373.4     73.4
Process Creation                                126.0      13113.2   1040.7
Shell Scripts (1 concurrent)                     42.4       7997.6   1886.2
Shell Scripts (16 concurrent)                     ---        901.5      ---
Shell Scripts (8 concurrent)                      6.0       1756.9   2928.1
System Call Overhead                          15000.0    1211587.2    807.7
                                                                   ========
System Benchmarks Index Score (Partial Only)                          916.9

------------------------------------------------------------------------
Benchmark Run: Sun Jan 23 2011 22:09:32 - 22:33:38
2 CPUs in system; running 2 parallel copies of tests

Dhrystone 2 using register variables       37996013.0 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     7678.6 MWIPS (9.7 s, 7 samples)
Execl Throughput                               8655.7 lps   (29.9 s, 2 samples)
Pipe Throughput                             2863704.1 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                 739539.5 lps   (10.0 s, 7 samples)
Process Creation                              25690.5 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                  14271.1 lpm   (60.0 s, 2 samples)
Shell Scripts (16 concurrent)                  1001.8 lpm   (60.1 s, 2 samples)
Shell Scripts (8 concurrent)                   1990.0 lpm   (60.0 s, 2 samples)
System Call Overhead                        2279039.9 lps   (10.0 s, 7 samples)

System Benchmarks Partial Index              BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   37996013.0   3255.9
Double-Precision Whetstone                       55.0       7678.6   1396.1
Execl Throughput                                 43.0       8655.7   2012.9
Pipe Throughput                               12440.0    2863704.1   2302.0
Pipe-based Context Switching                   4000.0     739539.5   1848.8
Process Creation                                126.0      25690.5   2038.9
Shell Scripts (1 concurrent)                     42.4      14271.1   3365.8
Shell Scripts (16 concurrent)                     ---       1001.8      ---
Shell Scripts (8 concurrent)                      6.0       1990.0   3316.7
System Call Overhead                          15000.0    2279039.9   1519.4
                                                                   ========
System Benchmarks Index Score (Partial Only)                         2226.9
 
Last edited:

eva2000

Well-Known Member
#2
With some help from Litespeed folks, got litespeed's inbuilt cache feature setup to work with vBulletin's boost/varnish plugin - basically just import/install product-boostv1.xml‎ on your vB 3.8.x or vB 4.0.x forums and setup Litespeed cache options in Litespeed admin console and drop in some vB specific .htaccess rewrite code and you're set.

The speed boost for guest visitors is crazy - between >16x to 123x times faster in different apachebench config tests!

Now to retest apachebench with above same settings for requests and concurrency with Litespeed inbuilt cache enabled resulted in Litespeed inbuilt cached tests being up to 721x times faster than Apache, or 91x times faster than Litespeed alone without the inbuilt cache enabled! :eek: Look at the cpu load for cached tests and no php memory usage at all :)

:: Jan 28th, 2011: Correction to findings
I have to revise my findings a bit for Nginx. The failed requests seem to be from content length varying between apachebench requests and not due to failure requests related to connections/exceptions errors. Content length varies due to dynamic nature of php pages tested as as long as content length reported by apachebench as Document Length is nearly the same, it still means a successfully served request regardless. Full summary here


Litespeed admin console cache policy and settings

* need to create cache path directory /lscache

Code:
Cache Storage Settings

Storage Path: /lscache
Max Object Size: 262144

Enable Cache:No
Cache Request with Query String:Yes
Cache Request with Cookie:Yes
Cache Response with Cookie:Yes
Ignore Request Cache-Control:Yes
Ignore Response Cache-Control:Yes
.htaccess used provided vB forum installed in domain/doc root

Code:
RewriteEngine on

RewriteRule test.php - [E=Cache-Control:max-age=45]
RewriteRule hello.php - [E=Cache-Control:max-age=45]

#RewriteCond %{HTTP_HOST} !^www\.domain\.com$
#RewriteRule /forum/(.*)$ http://www.domain.com/forum/$1 [R=301,L]


RewriteCond %{HTTP_COOKIE} !zzimloggedin=yes
RewriteCond %{HTTP_COOKIE} !zzuserid=
RewriteCond %{HTTP_COOKIE} !zzpassword=
RewriteCond %{REQUEST_METHOD} ^HEAD|GET$
RewriteCond %{QUERY_STRING} !s=[a-fA-F0-9]{32}
RewriteCond %{QUERY_STRING} !product=vbnexus
RewriteCond %{REQUEST_URI} !^(login|register|usercp|private|profile|cron|image|search)\.php$
RewriteCond %{REQUEST_URI} !^admincp
RewriteRule ^(.*\.php)?$ - [L,E=Cache-Control:max-age=120]
To confirm Litespeed Cache is working check the file header for X-LiteSpeed-Cache = hit

for index.php header via Firefox Firebug addon

Code:
Content-Encoding	gzip
Vary	Accept-Encoding
Date	Wed, 26 Jan 2011 16:40:00 GMT
Server	LiteSpeed
Connection	Keep-Alive
Keep-Alive	timeout=5, max=100
[B]X-LiteSpeed-Cache	hit[/B]
Content-Length	6406
X-Powered-By	PHP/5.3.5
Content-Type	text/html; charset=ISO-8859-1
Cache-Control	private
Pragma	private
Check /lscache directory contents

Code:
 ls -alhR /lscache
/lscache:
total 48K
drwxrwxrwx 11 nobody nobody 4.0K Jan 27 02:39 .
drwxr-xr-x 28 root   root   4.0K Jan 26 23:28 ..
drwx------  3 nobody nobody 4.0K Jan 27 01:45 0
drwx------  3 nobody nobody 4.0K Jan 27 01:44 1
drwx------  3 nobody nobody 4.0K Jan 27 01:35 2
drwx------  3 nobody nobody 4.0K Jan 27 01:39 6
drwx------  3 nobody nobody 4.0K Jan 27 01:45 7
drwx------  3 nobody nobody 4.0K Jan 27 01:45 9
drwx------  3 nobody nobody 4.0K Jan 27 01:39 a
drwx------  3 nobody nobody 4.0K Jan 27 01:45 b
drwx------  3 nobody nobody 4.0K Jan 27 02:39 e

/lscache/0:
total 12K
drwx------  3 nobody nobody 4.0K Jan 27 01:45 .
drwxrwxrwx 11 nobody nobody 4.0K Jan 27 02:39 ..
drwx------  3 nobody nobody 4.0K Jan 27 01:45 9

/lscache/0/9:
total 12K
drwx------ 3 nobody nobody 4.0K Jan 27 01:45 .
drwx------ 3 nobody nobody 4.0K Jan 27 01:45 ..
drwx------ 2 nobody nobody 4.0K Jan 27 01:45 8

/lscache/0/9/8:
total 20K
drwx------ 2 nobody nobody 4.0K Jan 27 01:45 .
drwx------ 3 nobody nobody 4.0K Jan 27 01:45 ..
-rw------- 1 nobody nobody 9.9K Jan 27 01:45 0987cec6

/lscache/1:
total 12K
drwx------  3 nobody nobody 4.0K Jan 27 01:44 .
drwxrwxrwx 11 nobody nobody 4.0K Jan 27 02:39 ..
drwx------  3 nobody nobody 4.0K Jan 27 01:44 a

/lscache/1/a:
total 12K
drwx------ 3 nobody nobody 4.0K Jan 27 01:44 .
drwx------ 3 nobody nobody 4.0K Jan 27 01:44 ..
drwx------ 2 nobody nobody 4.0K Jan 27 01:44 9

/lscache/1/a/9:
total 12K
drwx------ 2 nobody nobody 4.0K Jan 27 01:44 .
drwx------ 3 nobody nobody 4.0K Jan 27 01:44 ..
-rw------- 1 nobody nobody  201 Jan 27 01:44 1a977843

/lscache/2:
total 12K
drwx------  3 nobody nobody 4.0K Jan 27 01:35 .
drwxrwxrwx 11 nobody nobody 4.0K Jan 27 02:39 ..
drwx------  3 nobody nobody 4.0K Jan 27 01:35 9

/lscache/2/9:
total 12K
drwx------ 3 nobody nobody 4.0K Jan 27 01:35 .
drwx------ 3 nobody nobody 4.0K Jan 27 01:35 ..
drwx------ 2 nobody nobody 4.0K Jan 27 02:39 c

/lscache/2/9/c:
total 16K
drwx------ 2 nobody nobody 4.0K Jan 27 02:39 .
drwx------ 3 nobody nobody 4.0K Jan 27 01:35 ..
-rw------- 1 nobody nobody 6.5K Jan 27 02:39 29c52684

/lscache/6:
total 12K
drwx------  3 nobody nobody 4.0K Jan 27 01:39 .
drwxrwxrwx 11 nobody nobody 4.0K Jan 27 02:39 ..
drwx------  3 nobody nobody 4.0K Jan 27 01:39 4

/lscache/6/4:
total 12K
drwx------ 3 nobody nobody 4.0K Jan 27 01:39 .
drwx------ 3 nobody nobody 4.0K Jan 27 01:39 ..
drwx------ 2 nobody nobody 4.0K Jan 27 01:46 8

/lscache/6/4/8:
total 36K
drwx------ 2 nobody nobody 4.0K Jan 27 01:46 .
drwx------ 3 nobody nobody 4.0K Jan 27 01:39 ..
-rw------- 1 nobody nobody  25K Jan 27 01:46 6483cafa

/lscache/7:
total 12K
drwx------  3 nobody nobody 4.0K Jan 27 01:45 .
drwxrwxrwx 11 nobody nobody 4.0K Jan 27 02:39 ..
drwx------  3 nobody nobody 4.0K Jan 27 01:45 1

/lscache/7/1:
total 12K
drwx------ 3 nobody nobody 4.0K Jan 27 01:45 .
drwx------ 3 nobody nobody 4.0K Jan 27 01:45 ..
drwx------ 2 nobody nobody 4.0K Jan 27 01:45 4

/lscache/7/1/4:
total 16K
drwx------ 2 nobody nobody 4.0K Jan 27 01:45 .
drwx------ 3 nobody nobody 4.0K Jan 27 01:45 ..
-rw------- 1 nobody nobody 7.0K Jan 27 01:45 714e2f9d

/lscache/9:
total 12K
drwx------  3 nobody nobody 4.0K Jan 27 01:45 .
drwxrwxrwx 11 nobody nobody 4.0K Jan 27 02:39 ..
drwx------  3 nobody nobody 4.0K Jan 27 01:45 4

/lscache/9/4:
total 12K
drwx------ 3 nobody nobody 4.0K Jan 27 01:45 .
drwx------ 3 nobody nobody 4.0K Jan 27 01:45 ..
drwx------ 2 nobody nobody 4.0K Jan 27 01:45 a

/lscache/9/4/a:
total 20K
drwx------ 2 nobody nobody 4.0K Jan 27 01:45 .
drwx------ 3 nobody nobody 4.0K Jan 27 01:45 ..
-rw------- 1 nobody nobody  11K Jan 27 01:45 94a971a0

/lscache/a:
total 12K
drwx------  3 nobody nobody 4.0K Jan 27 01:39 .
drwxrwxrwx 11 nobody nobody 4.0K Jan 27 02:39 ..
drwx------  3 nobody nobody 4.0K Jan 27 01:39 e

/lscache/a/e:
total 12K
drwx------ 3 nobody nobody 4.0K Jan 27 01:39 .
drwx------ 3 nobody nobody 4.0K Jan 27 01:39 ..
drwx------ 2 nobody nobody 4.0K Jan 27 01:39 9

/lscache/a/e/9:
total 20K
drwx------ 2 nobody nobody 4.0K Jan 27 01:39 .
drwx------ 3 nobody nobody 4.0K Jan 27 01:39 ..
-rw------- 1 nobody nobody 8.2K Jan 27 01:39 ae962e7b

/lscache/b:
total 12K
drwx------  3 nobody nobody 4.0K Jan 27 01:45 .
drwxrwxrwx 11 nobody nobody 4.0K Jan 27 02:39 ..
drwx------  3 nobody nobody 4.0K Jan 27 01:45 8

/lscache/b/8:
total 12K
drwx------ 3 nobody nobody 4.0K Jan 27 01:45 .
drwx------ 3 nobody nobody 4.0K Jan 27 01:45 ..
drwx------ 2 nobody nobody 4.0K Jan 27 01:45 d

/lscache/b/8/d:
total 16K
drwx------ 2 nobody nobody 4.0K Jan 27 01:45 .
drwx------ 3 nobody nobody 4.0K Jan 27 01:45 ..
-rw------- 1 nobody nobody 5.9K Jan 27 01:45 b8d74ed3

/lscache/e:
total 12K
drwx------  3 nobody nobody 4.0K Jan 27 02:39 .
drwxrwxrwx 11 nobody nobody 4.0K Jan 27 02:39 ..
drwx------  3 nobody nobody 4.0K Jan 27 02:39 0

/lscache/e/0:
total 12K
drwx------ 3 nobody nobody 4.0K Jan 27 02:39 .
drwx------ 3 nobody nobody 4.0K Jan 27 02:39 ..
drwx------ 2 nobody nobody 4.0K Jan 27 02:39 d

/lscache/e/0/d:
total 12K
drwx------ 2 nobody nobody 4.0K Jan 27 02:39 .
drwx------ 3 nobody nobody 4.0K Jan 27 02:39 ..
-rw------- 1 nobody nobody  201 Jan 27 02:39 e0ddf880
Note: To be able to use Litespeed Inbuilt Cache feature you need >2 cpu Litespeed license. VPS and 1 CPU Licenses don't have the Inbuilt Cache supported.
 
Last edited:

eva2000

Well-Known Member
#3
Torture testing Litespeed Inbuilt cache against vB 3.8.6 PL1 front index.php page with apachebench at:

  • 5000 concurrency level + 1 million requests (virtualized server allocated 1GB memory) - cache timeout was set at 120 seconds but test run lasted 127.45 seconds, so the tail end caught some php load. PHP averaged around 10MB per process.
  • 5000 concurrency level + 1.5 million requests (virtualized server allocated 2GB memory) - cache timeout was set at 480 seconds while test ran only for 163.91 seconds, so well within cache timeout so no PHP usage.

Try the above with apache and only 1-2GB memory :D

Not a single failed request !

5000 concurrency + 1 million requests
Apachebench = 7,846.05 rps - over 240x times apache's performance handling 100x times the number of requests and 25x times the concurrency


Code:
ab -k -n 1000000 -c 5000 http://192.168.56.101/index.php
This is ApacheBench, Version 2.0.40-dev <$Revision: 1.146 $> apache-2.0
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Copyright 2006 The Apache Software Foundation, http://www.apache.org/

Benchmarking 192.168.56.101 (be patient)
Completed 100000 requests
Completed 200000 requests
Completed 300000 requests
Completed 400000 requests
Completed 500000 requests
Completed 600000 requests
Completed 700000 requests
Completed 800000 requests
Completed 900000 requests
Finished 1000000 requests


Server Software:        LiteSpeed
Server Hostname:        192.168.56.101
Server Port:            80

Document Path:          /index.php
Document Length:        19841 bytes

Concurrency Level:      5000
Time taken for tests:   127.452598 seconds
Complete requests:      1000000
Failed requests:        0
Write errors:           0
Keep-Alive requests:    418171
Total transferred:      20142500003 bytes
HTML transferred:       19874423323 bytes
Requests per second:    7846.05 [#/sec] (mean)
Time per request:       637.263 [ms] (mean)
Time per request:       0.127 [ms] (mean, across all concurrent requests)
Transfer rate:          154335.11 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        0   48  66.1     22     800
Processing:    89  571 694.6    567   21904
Waiting:       10  208 680.4    171   21283
Total:        109  619 702.5    618   21904

Percentage of the requests served within a certain time (ms)
  50%    618
  66%    665
  75%    690
  80%    705
  90%    756
  95%    871
  98%   1066
  99%   1191
 100%  21904 (longest request)
CPU Load

Code:
top - 16:27:02 up 23 min,  3 users,  load average: 9.43, 3.15, 1.56
Tasks: 136 total,   2 running, 134 sleeping,   0 stopped,   0 zombie
Cpu(s): 29.8%us, 15.0%sy,  5.0%ni, 19.3%id, 29.2%wa,  0.3%hi,  1.3%si,  0.0%st
Mem:   1026824k total,  1015532k used,    11292k free,      444k buffers
Swap:  2064376k total,   667332k used,  1397044k free,    30628k cached
5000 concurrency + 1.5 million requests​
Apachebench = 9,151.34 rps - 280x times apache's performance with 150x times number of requests and 25x times the concurrency.

Code:
ab -k -n 1500000 -c 5000 http://192.168.56.101/index.php
This is ApacheBench, Version 2.0.40-dev <$Revision: 1.146 $> apache-2.0
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Copyright 2006 The Apache Software Foundation, http://www.apache.org/

Benchmarking 192.168.56.101 (be patient)
Completed 150000 requests
Completed 300000 requests
Completed 450000 requests
Completed 600000 requests
Completed 750000 requests
Completed 900000 requests
Completed 1050000 requests
Completed 1200000 requests
Completed 1350000 requests
Finished 1500000 requests


Server Software:        LiteSpeed
Server Hostname:        192.168.56.101
Server Port:            80

Document Path:          /index.php
Document Length:        19841 bytes

Concurrency Level:      5000
Time taken for tests:   163.910341 seconds
Complete requests:      1500000
Failed requests:        0
Write errors:           0
Keep-Alive requests:    657470
Total transferred:      30208037959 bytes
HTML transferred:       29804908545 bytes
Requests per second:    9151.34 [#/sec] (mean)
Time per request:       546.368 [ms] (mean)
Time per request:       0.109 [ms] (mean, across all concurrent requests)
Transfer rate:          179976.66 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        0   42  53.5     17     309
Processing:   121  501 114.0    550     864
Waiting:       12  154  62.7    156     534
Total:        134  544 148.4    597    1016

Percentage of the requests served within a certain time (ms)
  50%    597
  66%    645
  75%    666
  80%    678
  90%    708
  95%    730
  98%    757
  99%    784
 100%   1016 (longest request)
CPU Load

Code:
top - 17:42:28 up  1:02,  3 users,  load average: 2.86, 1.33, 0.50
Tasks: 103 total,   4 running,  99 sleeping,   0 stopped,   0 zombie
Cpu0  : 29.5%us, 18.2%sy,  0.0%ni, 19.5%id,  4.0%wa,  0.0%hi, 28.8%si,  0.0%st
Cpu1  : 29.6%us, 14.3%sy,  0.0%ni, 35.5%id,  2.0%wa,  0.7%hi, 17.9%si,  0.0%st
Mem:   2059524k total,  2037892k used,    21632k free,      184k buffers
Swap:  2064376k total,    92112k used,  1972264k free,    38816k cached
 
Last edited:

eva2000

Well-Known Member
#4
:: Jan 28th, 2011: Correction to findings
I have to revise my findings a bit for Nginx in above tests. The failed requests seem to be from content length varying between apachebench requests and not due to failure requests related to connections/exceptions errors. Content length varies due to dynamic nature of php pages tested as as long as content length reported by apachebench as Document Length is nearly the same, it still means a successfully served request regardless.

However, not sure why litespeed did have any of these errors vBulletin index.php page tests. The reason I came across this is because I have been able to implement same litespeed cache method for use on wordpress as well. However, litespeed apachebench testing wordpress pages results in same failed request due to content length varying between requests. So I guess as long as apachebench Document Length is varying only slightly in size, you can ignore the failed requests if they are due to content length as reality is the php page would have been served.

Example: testing wordpress page non-cached nginx, litespeed and apache with only apachebench testing 2 requests and 1 concurrent all reported failed request due to content/document length

Failed request due to content/document length

Code:
Failed requests:        1
   (Connect: 0, Length: 1, Exceptions: 0)
Code:
ab -k -n 2 -c 1 http://192.168.56.101/wp/?p=1417
This is ApacheBench, Version 2.0.40-dev <$Revision: 1.146 $> apache-2.0
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Copyright 2006 The Apache Software Foundation, http://www.apache.org/

Benchmarking 192.168.56.101 (be patient).....done


Server Software:        nginx
Server Hostname:        192.168.56.101
Server Port:            80

Document Path:          /wp/?p=1417
Document Length:        39934 bytes

Concurrency Level:      1
Time taken for tests:   0.186644 seconds
Complete requests:      2
Failed requests:        1
   (Connect: 0, Length: 1, Exceptions: 0)
Write errors:           0
Keep-Alive requests:    0
Total transferred:      80258 bytes
HTML transferred:       79804 bytes
Requests per second:    10.72 [#/sec] (mean)
Time per request:       93.322 [ms] (mean)
Time per request:       93.322 [ms] (mean, across all concurrent requests)
Transfer rate:          417.91 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        0    0   0.0      0       0
Processing:    92   93   1.4     94      94
Waiting:       91   92   2.2     94      94
Total:         92   93   1.4     94      94

Percentage of the requests served within a certain time (ms)
  50%     94
  66%     94
  75%     94
  80%     94
  90%     94
  95%     94
  98%     94
  99%     94
 100%     94 (longest request)
Code:
ab -k -n 2 -c 1 http://192.168.56.101/wp/?p=1417
This is ApacheBench, Version 2.0.40-dev <$Revision: 1.146 $> apache-2.0
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Copyright 2006 The Apache Software Foundation, http://www.apache.org/

Benchmarking 192.168.56.101 (be patient).....done


Server Software:        LiteSpeed
Server Hostname:        192.168.56.101
Server Port:            80

Document Path:          /wp/?p=1417
Document Length:        39739 bytes

Concurrency Level:      1
Time taken for tests:   0.157730 seconds
Complete requests:      2
Failed requests:        1
   (Connect: 0, Length: 1, Exceptions: 0)
Write errors:           0
Keep-Alive requests:    0
Total transferred:      80100 bytes
HTML transferred:       79648 bytes
Requests per second:    12.68 [#/sec] (mean)
Time per request:       78.865 [ms] (mean)
Time per request:       78.865 [ms] (mean, across all concurrent requests)
Transfer rate:          494.52 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        0    0   0.0      0       0
Processing:    78   78   0.0     78      78
Waiting:       68   68   0.0     68      68
Total:         78   78   0.0     78      78

Percentage of the requests served within a certain time (ms)
  50%     78
  66%     78
  75%     78
  80%     78
  90%     78
  95%     78
  98%     78
  99%     78
 100%     78 (longest request)
Code:
ab -k -n 2 -c 1 http://192.168.56.101/wp/?p=1417
This is ApacheBench, Version 2.0.40-dev <$Revision: 1.146 $> apache-2.0
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Copyright 2006 The Apache Software Foundation, http://www.apache.org/

Benchmarking 192.168.56.101 (be patient).....done


Server Software:        Apache/2.2.3
Server Hostname:        192.168.56.101
Server Port:            80

Document Path:          /wp/?p=1417
Document Length:        39772 bytes

Concurrency Level:      1
Time taken for tests:   0.188623 seconds
Complete requests:      2
Failed requests:        1
   (Connect: 0, Length: 1, Exceptions: 0)
Write errors:           0
Keep-Alive requests:    0
Total transferred:      80605 bytes
HTML transferred:       80097 bytes
Requests per second:    10.60 [#/sec] (mean)
Time per request:       94.312 [ms] (mean)
Time per request:       94.312 [ms] (mean, across all concurrent requests)
Transfer rate:          413.52 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        0    0   0.0      0       0
Processing:    94   94   0.0     94      94
Waiting:       49   49   1.0     50      50
Total:         94   94   0.0     94      94

Percentage of the requests served within a certain time (ms)
  50%     94
  66%     94
  75%     94
  80%     94
  90%     94
  95%     94
  98%     94
  99%     94
 100%     94 (longest request)
 
Last edited:

eva2000

Well-Known Member
#6
Very nice Benchmark.
Still can not make Guest Cache working with VBSEO.
Thanks. Looks like after wordpress + litespeed cache, vB + vBSEO + litespeed cache would be next on my list of things to conquer as alot of the private clients including myself use vBSEO with vB.

I finally managed to get Varnish VCL tuned to work with vBulletin 3.8.6 PL1 as well, so updated the above table to include Apache + Varnish 2.1.5 vB index.php tests to the list as well as update the table column to remove failed requests in the calculation of adjusted rps for nginx.



Looks like when you include mysql backend as a factor in testing vB index.php page, litespeed cache just edges out Apache + Varnish. Unlike, with php based test.php file where Apache + Varnish took the lead - although it seems i hit a litespeed cache memory leak bug in those tests.
 
Last edited:

fastproxy

Well-Known Member
#8
Thanks for the nice guide,

i got the cache working, i have a server using haproxy/nginx as load balancer before litespeed webserver.
Code:
Server	nginx
Date	Tue, 15 Feb 2011 19:08:16 GMT
Content-Type	text/html; charset=utf-8
Connection	keep-alive
Content-Encoding	gzip
Vary	Accept-Encoding, User-Agent
X-LiteSpeed-Cache	hit
Content-Length	22072
X-Powered-By	PHP/5.2.14
Cache-Control	private
Pragma	private
X-UA-Compatible	IE=7
but some how some pages show i logged in as someone else account?look like the page i was loading belong to someone else.

Is there something missing?
I am using VBB 3.8.6


Update:

I found the problem, it was the .htaccess i copy from your but forgot to edit according to my server:
RewriteCond %{HTTP_COOKIE} !zzimloggedin=yes
RewriteCond %{HTTP_COOKIE} !zzuserid=
RewriteCond %{HTTP_COOKIE} !zzpassword=
Thanks
 
Last edited:
Top