[Performance Testing Tool]-http_load

Advantage: few parameters; disadvantage:Url.txtHigh file requirements, not as goodSiegeConvenient

Http_load is developed by the Acme lab. This tool runs concurrently to test the Web server throughput and load. Unlike other stress testing tools, the tool runs in a single process and generally does not shut down the client. In UNIX, besides HTTP, it can also be used to test HTTPS requests. Official Address: http://www.acme.com/software/http_load/

 

Installation:
Tar zxvf http_load-12mar2006.tar
CD http_load-12mar2006
There are spaces between make & make install/* makeinstall, and make install requires the root permission */

 

Parameter description:
Xxxxxxxxxxxxx :~ Http_load-12mar2006 $ http_load-help
Usage: http_load [-checksum] [-Throttle] [-proxy host: Port] [-verbose] [-Timeout secs] [-Sip sip_file]
-Parallel n |-rate N [-jitter]

-Fetches n |-seconds n

Url_file/* the URL text file to be accessed. The file format is one URL per line. It is better to have more than 50-URLs for better results */
One start specifier, either-PArallel or-RAte, is required./* select either-parallel or-rate */
One end specifier, either-FEtches or-SEconds, is required./* select either-feches or-Second */

 

-Abbreviated as parallel-P: Number of concurrent user processes/* http_load-P500-S 300 user.txt, 500 concurrent processes, 300 seconds */

-Rate abbreviation-R: The Access frequency per second/* For example, http_load-R 5-S 10 url.txt, which means that the Access frequency per second is 5 times, and a total of 10 seconds */

-Short for fetches-F: Total number of visits/* The use of fetches allows the client to ensure that all requests are processed, making it easier for the tested server to be under pressure */

-Seconds-S: Total access time

 

Basic usage:

Xxxxxxxxxxxxx :~ /Http_load-12mar2006 $ http_load-P 10-S 5 urls.txt/* concurrent 10 processes, access 5 seconds. The file format in urls.txt is one URL per line. */

 

Operation Result:(TestedUrl.txtAre:Http://www.test.com/index.html)

13995 fetches, 10max parallel, 21948 bytes, in 5 seconds/* run a total of 13995 requests. The maximum number of concurrent processes is 10, and the total data transmitted is 21948 bytes, running time: 5 seconds */

1.56827 mean Bytes/connection/* 21948 Bytes/13995 requests = the average data volume transmitted by each connection is 1.56827 bytes */

2799 fetches/sec, 4389.6 Bytes/SEC/* The number of queries that the server can respond to per second is 2799 (13995/5), and the amount of data transmitted per second is 4389.6 bytes */

Msecs/Connect: 1.74836 mean, 3.757 Max, 1.439 min/* The average access time for each connection is 1.748 Ms */

Msecs/first-response: 1.79729 mean, 5.463 Max, 1.484 min /**/

HTTP responsecodes:

Code 200 -- 236

 

Discussion/Question:

 

• What is the difference between msecs/connect and msecs/first-response in the test results?