- Sign on to the IBM i system with a user profile that has *ALLOBJ and *SECADM special authorities.
- Display the current value of QALWOBJRST and then change it to *ALL if needed.
- Execute the following CL command to display the current value of QALWOBJRST:
- If the current value is NOT *ALL, record the current value (so you remember what to change it back to) and then change the value to *ALL using the following CL command:
CHGSYSVAL SYSVAL(QALWOBJRST) VALUE(*ALL)
- The WebSphere Group PTF has also provided a local repository for the latest IBM Installation Manager (IM) fix pack, and it is recommended that the IM be updated before you install the WebSphere fix packs provided by the WebSphere Group PTF. The Readme.txt and Readme.html files for the IM fix pack can be found here:
- Stop all servers and applications on the WebSphere Application Server installation that is being updated. These processes may include application server processes, nodeagent processes, deployment manager processes, job manager processes and WebSphere Application Server client applications.
This is most easily done by ending all of the application server JVM jobs and the subsystem for the WAS installation version and edition you are updating.
For example, if I am updating my WAS v8.0 Express installation; I would end all of the WAS v8.0 Express jobs in the QWAS8 subsystem first using the stopServer Qshell command for each profile job currently active in the subsystem:
STRQSH cd <product_location>/bin ./stopServer <server> -profileName <profile>
Note: The <server> and <profile> values are case sensitive and must match the names exactly. Then end the QWAS8 subsystem using the CL command:
ENDSBS QWAS8 *IMMED
Install the fix pack.
From a system command line, invoke the following SBMJOB command to submit the installation process to batch via the UDPATEWAS job; where
_location>, <offering_ID>, <product_location> and <was_fixpack_location> are defined above:
SBMJOB CMD(STRQSH CMD('umask 022;<Installation_Manager
_location>/eclipse/tools/imcl install <offering_ID> -acceptLicense -installationDirectory <product_location> -repositories <was_fixpack_location>')) JOB(UPDATEWAS) ALWMLTTHD(*YES) JOBQ(QSYS/QSYSNOMAX)
Note: IBM recommends the use of the SBMJOB command to submit the installation to batch instead of running the command interactively. Running the installation process in batch will prevent issues from occurring if your interactive session loses connectivity or ends unexpectedly.
Optionally, the IM imcl command can be executed interactively using the following steps:
- On a CL command line, run the STRQSH command to start the Qshell command shell.
- Make sure that the umask is set to 022.
- To verify the umask setting, issue the following command:
- To set the umask setting to 022, issue the following command:
- Change to the <Installation_Manager_
location>/eclipse/tools directory by executing the following command:
- Execute the fix pack install process
./imcl install <offering_ID> -acceptLicense -installationDirectory <product_location> -repositories <was_fixpack_location> -showVerboseProgress
Note: Installation of the fix pack may take over an hour depending on your IBM i system. If memory available to the IBM Installation Manager is limited, the update can take much longer, potentially several hours. You will know the installation process has completed when the UPDATEWAS job ends or the "./imcl install" command returns to a dollar sign in Qshell.
- To verifiy that the fix pack correctly updated the WebSphere installation, execute the following Qshell command:
The 'Version' that this command lists should be the same as the fix pack version just installed.
- End and restart the HTTP Admin instance if it was active during the update. To end and restart the HTTP Admin invoke the following commands from a system command line:
ENDTCPSVR SERVER(*HTTP) HTTPSVR(*ADMIN)
STRTCPSVR SERVER(*HTTP) HTTPSVR(*ADMIN)
Note: In addition, you will want to restart any HTTP server instances associated with your WAS v8.0 application servers. This can be done using the following commands.
ENDTCPSVR SERVER(*HTTP) HTTPSVR(<HTTPserver>)
STRTCPSVR SERVER(*HTTP) HTTPSVR(<HTTPserver>)
- Once the IBM WAS Fix Pack installation has been verified as successful, you are free to start your IBM WebSphere Application Server profiles back up using the following Qshell command.
<product_location>/bin/startServer <server> -profileName <profile>
In addition, you also may want to change your QALWOBJRST IBM i system value back to its original value using the following CL command:
CHGSYSVAL SYSVAL(QALWOBJRST) VALUE(<previousValue>)
- The fix pack installation process is now complete. If you experience issues or have questions along the way, please contact the IBM i Global Support Center at 1-800-IBM-SERV and open a new service request
Note: The <was_fixpack_location> directories can be large and are not managed or deleted by WebSphere. After all WebSphere installations have been updated and they are no longer needed, the system administrator can delete them and all their contents.
Tip: IBM recommends saving the previous fix pack installation repository in case you need to rollback the fix pack to the previous version.
Additional information about installing a fix pack may be found in the information center article, "Installing fix packs on IBM i operating systems using the Installation Manager command line", for Express, Base, and Network Deployment.