Configuring the KHU module for HTTP Server monitoring

The KHU module gets traditional resource and performance metrics. KHU is the agent code (unique code among all agents) for HTTP Server agent and the KHU module talks with HTTP Server agent process.

Before you begin

It is strongly suggested that install the HTTP Server agent when HTTP Server process is running because the agent installation can discover the running HTTP Server process and generate needed reference file.

There are two files involved in the configuration of the HTTP Server agent:

HTTP Server configuration file: It comes with every HTTP server.
HTTP Server agent data collector configuration file: It is also called reference file that is generated by HTTP Server agent when it discovers the running HTTP Server.

What you need to do is to include the reference file in the HTTP Server configuration file. And then restart HTTP Server. For details, see the Procedure section.

The HTTP Server configuration file

Each HTTP server has a configuration file by default: http_server_install_dir/conf/httpd.conf. In some environments, this file name might be customized. Check the exact file name with the HTTP Server administrator.

The HTTP Server agent data collector configuration file

When the HTTP Server agent discovers the HTTP server, it generates a data collector configuration file in the install_dir/tmp/khu directory.

If multiple HTTP Servers are discovered, multiple corresponding HTTP Server agent configuration files get generated.

The HTTP Server agent configuration file name has the format: khu.http_server_install_dir.conf.httpd.conf
The first part of the agent configuration file name is khu, in which khu is the HTTP server agent code.
The remaining part is the discovered HTTP server configuration file name with full path, in which / is replaced by . . For example, possible file names are as follows:

khu.usr.local.apache24.conf.httpd.conf
khu.C.Program Files.IBM.HTTPServer.conf.httpd.conf

The following elements are inside the HTTP Server agent data collector configuration file :

Details of the path of the httpd.conf file that the HTTP Server uses, for example, KhuShmemPath "/IBM/HTTPServer/conf/httpd.conf".
Location of the library to load
Permissions that are associated with the shared memory

For more information about the configuration file, see Reference: HTTP Server agent configuration file samples.

Note: Normally, you do not need to modify this configuration file. Just locate and record the location.

About this task

After you install the HTTP Server agent, enable the HTTP Server agent for data collection.

Note: You may also need to do this process on the following special occasions:

After you upgrade to HTTP Server agent version 1.0.0.4, the new alias causes the existing HTTP Server agent nodes to become offline in the Cloud APM console.
After you upgrade from version 1.0.0.4, the existing HTTP Server agent nodes can become offline in the Cloud APM console. This can occur if you have multiple HTTP server instances with similar agent configuration file names, for example, httpd and httpd01.

Important:

The netstat command-line network utility tool is needed for the HTTP Server agent to successfully discover the running HTTP server.
To solve the agent node offline problem that happens after the upgrade, you must add the new HTTP server instance in the Cloud APM console after you complete this task.

Procedure

To activate data collection, you must reference the data collector configuration file in the HTTP server configuration file by using the Include statement. Append the following statement to the end of the HTTP Server configuration file:

Include “agent_install_dir/tmp/khu/khu.http_server_install_dir.conf.httpd.conf“

For example,
If you have an IBM® HTTP Server that is installed in the /opt/IBM/HTTPServer directory and the data collector configuration file name with full path is like the following: /opt/ibm/apm/agent/tmp/khu/khu.opt.IBM.HTTPServer.conf.httpd.conf. Append the following statement to the /opt/IBM/HTTPServer/conf/httpd.conf HTTP server configuration file:
```
Include “/opt/ibm/apm/agent/tmp/khu/khu.opt.IBM.HTTPServer.conf.httpd.conf“
```
If you have an IBM® HTTP Server that is installed in the C:\ProgramFiles\IBM\HTTPServer directory and the data collector configuration file name with full path is like the following: C:\IBM\APM\tmp\khu\khu.C.Program Files.IBM.HTTPServer.conf.httpd.conf. Append the following statement to the C:\Program Files\IBM\HTTPServer\conf\httpd.conf HTTP server configuration file:
```
Include “C:\IBM\APM\tmp\khu\khu.C.Program Files.IBM.HTTPServer.conf.httpd.conf“
```
Change to the following directory:
HTTP_server_installation_directory/bin

Restart the HTTP Server. For example:

./apachectl -k stop
./apachectl -k start

httpd.exe -k stop
httpd.exe -k start

Results

You have successfully configured the agent.

What to do next

Now, you can verify the HTTP Server agent data is displayed in the Cloud APM console. For instructions on how to start the Cloud APM console, see Starting the Cloud APM console. For information about using the Applications editor, see Managing applications.

Note: If there is no traffic in the HTTP server, you will not see data in the Cloud APM console.