
The Installation Manager command line tool (imcl) has two different commands that can be used to install one or more packages.
- install
- input (this article)
When using the install command, all of the parameters must be used on the command line. When using the input command, most of the parameters are put in an XML file.
IMPORTANT
The imcl command can only be used if this is the first install of Installation Manager. If Installation Manager is already installed, the following will be displayed when attempting to install Installation Manager using the imcl command:
"Installation Manager is already installed"
The installc or userinstc commands can be used to install another installation of Installation Manager.
Install Installation Manager
If you have not yet installed IBMs Installation Manager, check out my articles:
- Install Installation Manager on Linux using the installc or userinstc command
- Install Installation Manager on Linux using the imcl install command
- Install Installation Manager on Linux using the imcl input command
Create the XML file
Create an XML file. Typically, the name of the XML file is response.xml. However, the file name does not have to be response.xml. It can be named anything you want, as long as it's an XML file.
response.xml
Or, instead of manually creating the response.xml file, you can create the response file by issuing the following command, which will launch the Installation Manager GUI, so this must be done in a graphical environment. Follow the steps on installing WebSphere using the GUI.
/opt/IBM/InstallationManager/eclipse/IBMIM
-record
/path/to/response.xml
-skipInstall
/opt/IBM
remote repository vs. local repository
In the XML file, you will need to include the location of the repository that contains the installation files for the product. One option is to use the IBM HTTP URL that contains the installation files for a product. This is known as an "IBM service repostiory". Additionally, when repository location in your XML points to an IBM HTTP URL (e.g. IBM service repository), you will need to provide your IBM username and password to perform the install. One way to provide your username and password is by using the -prompt, which will prompt you to provide your username and password. Another option is to store your username in a secure storage file and to store your password in a master password file, and to then using the -secureStorageFile and -masterPasswordFile options on the command line.
<server>
<repository location="http://www.ibm.com/software/repositorymanager/com.ibm.websphere.BASE.v90" />
</server>
Another option is to download the installation files from Fix Central or using the Packaging Utility, and to then specify the location of the downloaded installation files. This is known as a "local repository". An advantage to this approach is that there will be no need to provide your IBM username and password to perform the install, as the installation files reside on your server.
<server>
<repository location="/tmp/websphere" />
</server>
Here is an example of a complete response.xml file. The listAvailablePackages command can be used to get the offering id of the packages to install.
<?xml version="1.0" encoding="UTF-8"?>
<agent-input>
<variables>
<variable name='sharedLocation' value='/opt/IBM/IMShared'/>
</variables>
<server>
<repository location="http://www.ibm.com/software/repositorymanager/com.ibm.websphere.BASE.v90" />
</server>
<profile id='IBM WebSphere Application Server V9.0' installLocation='/opt/IBM/WebSpere/AppServer'>
<data key='user.import.profile' value='false'/>
<data key='cic.selector.nl' value='en'/>
</profile>
<install modify='false'>
<offering id='com.ibm.websphere.BASE.v90'
profile='IBM WebSphere Application Server V9.0'
features='core.feature,ejbdeploy,thinclient,embeddablecontainer' installFixes='none'/>
<offering id='com.ibm.java.jdk.v8'
profile='IBM WebSphere Application Server V9.0'
features='com.ibm.sdk.8'/>
</install>
</agent-input>
Perform the install
If repository location in your XML points to an IBM HTTP URL (e.g. IBM service repository), you will need to provide your IBM username and password to perform the install by using the -prompt or -secureStorageFile /path/to/secure_storage.txt and -masterPasswordFile /path/to/master_password.txt options on the command line.
/opt/IBM/InstallationManager/eclipse/tools/imcl
input
/opt/IBM/response.xml
-acceptLicense
-showProgress or -showVerboseProgress
Post installation
The websphere_installation_directory/logs/install/log.txt file can be reviewed to ensure the installation was successful.
The Installation Verification Tool can be used to ensure the installation was successful.
The listInstalledPackages command can be used to list installed packages.
The versionInfo.sh utility can be used to verify that the application server was successfully installed.
~]# /opt/IBM/WebSphere/AppServer/bin/versionInfo.sh
. . .
Installed Product
---------------------------------------------------
Name IBM WebSphere Application Server
Version 9.0.0.5
ID Base
. . .
After the application server is installed, you will create a profile.
Did you find this article helpful?
If so, consider buying me a coffee over at