Linux -Blue forest free software | Return to home page | Site Map | Search WWW | Contact Us |
Your current position : Homepage > Free Software > Technological exchanges >System Management


    

Blue Forest http://www.lslnet.com at 20:18 on April 6, 2006


[Help] server stress tests do FREEBSD

Today as a server installed, the system FREEBSD is primarily used for running Web server, APACHE+MYSQL+PHP also out, and now want to do something in this table machine server capacity stress tests to see how, but I do not know what should be the most accurate method to test, to ask, thank you.

[Help] server stress tests do FREEBSD

Ab

[Help] server stress tests do FREEBSD

I still find that I use the method has been tested, together with the results!
Supporting the Apache server software, called an ab (ApacheBench) tool. ApacheBench Apache server is mainly used to test the efficiency of the implementation, we used to CGI vs. ApacheBench. FastCGI the assessment tool.

ApacheBench simulation up to address the specific URL for online requests, can also simulate the same time points out the same month a request online, web site development and use ApacheBench during simulation can actually help us in the lane. simulation server setting or adjusting the data as the basis for a procedure. ApacheBench usage as follows :

Usage : /usr/local/apache/bin/ab [options] [http://]hostname[]/path
Options are :
N requests Number of requests to perform
Number of multiple requests to make concurrency Ones
T timelimit Seconds to max. Wait for responses
- P postfile File containg data to POST
T content-type Content-type header for POSTing
How much verbosity-V [to print troubleshooting
W Print out results in HTML tables
X attributes String to insert as table attributes
Tr y attributes String to insert as attributes
Z attributes String to insert as td or th attributes
Print version number and exit-V
K Use HTTP KeepAlive feature
- H Display usage information (this message)

If we are to do the test echo.fcg simulation 1000 times, the request line, but there are 20 parallel the same time (concurrent) line in the case, if implemented in line mode

$ 20 ab-n-c 1000 http://localhost/fcgi-bin/echo.fcg
Wait and the results will show up ApacheBench.

This is ApacheBench, Version 1.3
 Privacy Policy  Copyright © 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
 Privacy Policy  Copyright © 1998-1999 The Apache Group, http://www.apache.org/

Benchmarking localhost (be patient). . . Apache/1.3.6 Server Software :
Server : Hostname localhost
Server Port : 80
Document Path : /fcgi-bin/echo.fcg
Document Length : 995 bytes
Concurrency Level : 20
Time taken for tests : 6.859 seconds
Complete requests : 1000
Failed requests : 0
Total 1142000 bytes transferred :
HTML : 995,000 bytes transferred
Requests : 145.79 per second
Transfer rate : 166.50 kbps received

Connnection Times (ms)
Min avg max

Connect : 0 4 61
Processing : 62 128 771
Total : 62 132 832

These results point out that the request line at the same time 20 (Concurrency Level) circumstances, the completion of the on-line request for 1000, spent a total of 6.859 seconds (Time taken for tests). Therefore, the procedures can be handled per second on average (Requests per second) 145.79-line request.
In the next assessment test, we will be able to handle the line number of requests per second was to compare the effectiveness of the basis.

[Help] server stress tests do FREEBSD

My server the test results are as follows :

. N 1000-c 20 http://8899.kmip.net/ /ab
This is ApacheBench, Version 2.0.41-dev "$Revision : 1.121.2.12 $>; apache-2.0
Copyright (c) 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Copyright (c) 1998-2002 The Apache Software Foundation, http://www.apache.org/

Benchmarking 8899.kmip.net (be patient)
Completed 100 requests
Completed 200 requests
Completed 300 requests
Completed 400 requests
Completed 500 requests
Completed 600 requests
Completed 700 requests
Completed 800 requests
Completed 900 requests
Finished 1000 requests


Apache/2.0.52 Server Software :
Server : Hostname 8899.kmip.net
Server Port : 80

Document Path : /
Document Length : 24,813 bytes

Concurrency Level : 20
Time taken for tests : 206.74303 seconds
Complete requests : 1000
Failed requests : 591
(Connect : 0, Length : 591, Exceptions : 0)
Write errors : 0
Total : 25227965 bytes transferred
HTML : 24826163 bytes transferred
Requests : 485 [#/sec] per second (mean)
Time per request : 4121.486 [ms] (mean)
Time per request : 206.074 [ms] (mean, across all concurrent requests)
Transfer rate : 119.55 [Kbytes/sec] received

Connection Times (ms)
Mean[+/-sd] median max min
Connect : 134.9 0 3000 0 7
Processing : 4042 6654 2363 4077 21.55
Waiting : 1280 2253 3468 2400 4017
Total : 4044 7194 2364 4084 254.1

Percentage of the requests served within a certain time (ms)
50% 4044
66% 4094
75% 4135
80% 4174
90% 4236
95% 4292
98% 4599
99% 4935
100% 7194 (longest request)

[Help] server stress tests do FREEBSD

I can not remember what.

[Help] server stress tests do FREEBSD

Collection

[Help] server stress tests do FREEBSD

Here are my operations with the Windows virtual machine under the 5.3BSD VMWave
. 200 1000 n-c http://192.168.1.5/ /ab

This is ApacheBench, Version 2.0.41-dev "$Revision : 1.121.2.12 $>; apache-2.0
Copyright (c) 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Copyright (c) 1998-2002 The Apache Software Foundation, http://www.apache.org/

Benchmarking 192.168.1.5 (be patient)


Apache/2.0.54 Server Software :
Server : Hostname 192.168.1.5
Server Port : 80

Document Path : /
Document Length : 1456 bytes

Concurrency Level : 200
Time taken for tests : 7.985228 seconds
Complete requests : 1000
Failed requests : 0
Write errors : 0
Total 2031176 bytes transferred :
HTML : 1549184 bytes transferred
Requests : 125.23 [#/sec] per second (mean)
Time per request : 1597.046 [ms] (mean)
Time : 7.985 per request [ms] (mean, across all concurrent requests)
Transfer rate : 248.33 [Kbytes/sec] received

Connection Times (ms)
Mean[+/-sd] median max min
Connect : 9 518 1285.2 140 6450
Processing : 1314 981 1889 140 991
Waiting : 905 1584 291.7 92 880
Total : 1070 7482 321 1509 1337.1

Percentage of the requests served within a certain time (ms)
50% 1070
66% 1304
75% 1425
80% 1649
90% 2011
95% 4419
98% 7461
99% 7470
100% 7482 (longest request)

The results indicate that, at the same time request line 200 (Concurrency Level) circumstances, the completion of the on-line request for 1000, spent a total of 7.985228 seconds
(Time taken for tests), this procedure can handle per second on average (Requests per second) 248.33-line request.
In the next assessment test, we will be able to handle the line number of requests per second was to compare the effectiveness of the basis.

[Help] server stress tests do FREEBSD

These tests do find is a very important point of inflection point performance, so suddenly on 1000, despite the 1000 can tell you with "transportation" or "No", but which might mask the real issues. My suggestion is that up to 100,100 increase, and then see markedly slowed the response time, a time to measure CPU, network, which FS or a memory problem.

[Help] server stress tests do FREEBSD

Total : 25227965 bytes transferred
HTML : 24826163 bytes transferred
Requests : 485 [#/sec] per second (mean)
Time per request : 4121.486 [ms] (mean)
Time per request : 206.074 [ms] (mean, across all concurrent requests)
Transfer rate : 119.55 [Kbytes/sec] received

Lz how the data so weak?

[Help] server stress tests do FREEBSD

Why do I find this order, it needs to install?

[Help] server stress tests do FREEBSD

End x installed there, or they can do a cp.



 Privacy Policy  Copyright © 1999-2000 LSLNET.COM. All rights reserved. Blue Forest website owners. E-mail : Webmaster@lslnet.com