Use weblog Experet to analyze logs for performance requirements

First, download weblog expert log Analysis tool

Official website: http://www.weblogexpert.com/download.htm

Second, installation

Installation is simple, click "Next" Default installation can

Iii. configuring Tomcat to automatically generate access logs (Localhost_access_yyyy_mm_dd.log)

Tomcat does not turn on auto-generated access logs by default, that is, there are no access.log-like log files in the Tomcat/logs directory, and the Tomcat access log needs to be modified Tomcat/conf/server.xml

Mode one: Remove the following configuration comments from the Server.xml file, and change the pattern to combined

Mode Two: Add the following configuration directly in the Server.xml file

<className= "Org.apache.catalina.valves.AccessLogValve"               Directory  = "Logs"  prefix= "Localhost_access_log." suffix = ". txt"                pattern= "combined"  filedateformat= "Yyyy-mm-dd"  resolvehosts= "false"/>

Iv. using weblog expert tools to generate access log reports

1. Create a new project

Note: If you do not know the website address, you can write localhost

2. Select log file, file type, path

Note: Select Local log file here

3. Analysis time range Select all range of activities

4. Add a trace file (or do not add, click Next to)

5. Add a filter analysis log file

6. Specify the parse log report path to complete.

V. Analyze logs for performance requirements

1.General statistics-summary

Extract information:

Average Hits per day (average daily clicks): The larger the number of hits, the greater the pressure on the server.

Cached requests (Cache request): The number of cache requests is small, indicating that the system cache needs to be increased.

Average page views per day (average daily page visits): The more pages you visit, the greater the stress on the server performance.

Average Visitors per day (average daily visitors): Average daily visitors = Average daily number of people online.

Average Bandwidth per day (average daily bandwidth flow)/7200: This formula calculates the bandwidth per second at peak (KB/S), which is the peak throughput.

2.Activity Statistics

Daily Visitors: Peak count, peak concurrency = Peak *10%

Daily Hits: Peak values

Daily Bandwidth: Peak values

Activity by Hour of the day focuses on which time period, and how many hours it lasts

3.Access Statistics

Daily page Acces, most Popular pages: Determine performance test scenarios based on frequently visited pages.

Daily image access, most requested Images: Identify performance test points based on frequently accessed images, and can refer to these pictures to optimize front-end performance.

4.Errors

Daily error Types: View specific error codes and error request links and feedback to developers.

Use weblog Experet to analyze logs for performance requirements