Gather information with the collector tool
The collector tool gathers information about the WebSphere Application Server installation and packages it in a JAR file that we can send to IBM Customer Support to assist in determining and analyzing our problem. Information in the JAR file includes logs, property files, configuration files, operating system and Java data, and the presence and level of each software prerequisite.
The sort of information gathered is not something that most people use. In fact, the collector tool packages its output into a JAR file. IBM includes the collector tool in the product code, along with other tools that help capture the information that provide when reporting a problem.
There are two phases of using the collector tool. The first phase runs the collector tool on our WAS product and produces a JAR file. The IBM Support team performs the second phase, which is analyzing the JAR file that the collector program produces. The collector program runs to completion as it creates the JAR file, despite any errors that it might find like missing files or invalid commands. The collector tool collects as much data in the JAR file as possible.
The collector tool is a Java application that requires a Java SE Runtime Environment 6 (JRE6) to run.
The tool is within the installation root directory for WAS ND. But we run the tool from a working directory that we create outside of the installation root directory.
There are two ways to run the collector tool. Run the collector tool to collect summary data or to traverse the system to gather relevant files and command results. The collector tool produces a JAR file of information needed to determine and solve a problem. The collector summary option produces a lightweight collection of version and other information that is useful when first reporting the problem to IBM Support. Run the collector tool from the root user or from the administrator user to access system files that contain information about kernel settings, installed packages, and other vital data.
The tool collects information about the default profile if we do not use the optional parameter to identify another profile.
Run the collector tool
- Log on to the system as root or a member of the administrator group on a Windows platform.
- Verify that Java 1.2.2 or higher is available in the path.
The collector program requires Java code to run. It also collects data about the IBM Developer Kit, Java Technology Edition in which it runs.
If there are multiple Developer Kits on the system, verify that the one that the WAS product uses is the one in the path for the collector program.
If the Developer Kit being used by the WAS is not available, put another Developer Kit in the path for the collector program so that we can collect everything except data about the Developer Kit that WAS ND is using.
- Verify that all necessary information is in the path being used by the collector program and that we are not running the program from within the WAS product installation root directory.
- (UNIX) Verify that the path contains the following system directories:
- /bin
- /sbin
- /usr/bin
- /usr/sbin
- (Windows) Include regedit in the path.
- Make a working directory where we can start the collector program.
- Make the working directory the current directory.
- Run the collector program by entering the fully qualified command from the command line of the working directory.
app_server_root/bin/collector.sh
Use the command with no additional parameter to gather one copy of the profile data and data from each server in the node, and to store the data in a single JAR output file.
- To gather data from a specific profile that might not be the default profile.
app_server_root/bin/collector.sh -profileName profile
- Optional: We can run the collector tool from a profile's bin directory instead of the app_server_root/bin/ directory.
Run the following command...
profile_root/bin/collector.sh
We should get the same output if we run the collector tool from the bin directory of profile_root as we would running it from app_server_root.
Issuing the command from the profile also runs the setupCmdLine.bat/sh file in the profile's bin directory. This file sets an environment parameter that the collector uses to determine which profile's data to collect.
To run this command for the deployment manager, for example, issue the following at a prompt:
app_server_root/profiles/dmgr/bin/collector.sh
where dmgr is the profile name for the deployment manager.
The collector program creates the Collector.log log file and an output JAR file in the current directory.
The name of the JAR file is composed of the host name, cell name, node name, and profile name:
host-cell-node-profile.JAR
The Collector.log log file is one of the files collected in the jar file.
What to do next
Send the host_name-cell-node-profile.JAR file to IBM Support for analysis.
Subtopics