I have a puzzling (frustrating, baffling, pick an adjective) problem with Java 1.5.0_06. Believe me when I say I have searched and found no answer.

I tried to deploy the extracted MSI but the user who tested the install found that the Java Update registry key was not being set properly despite the command line variables (JAVAUPDATE=0, JU=0). So instead I set up the exe to deploy with a customized command line from the Sun 1.5.0 site.

jre-1_5_0_06-windows-i586-p.exe /S /v"/qn /l*v C:\logs\jre15006n.log ADDLOCAL=ALL REBOOT=R IEXPLORER=1 JAVAUPDATE=0 ALLUSERS=1 JU=0"
Our users already have Java 1.5.0_01 and when I tried to make 1.5.0_06 upgrade automatically it failed. So I set up a separate uninstall for the _01 package. It works well, the uninstall is very clean and there is a required reboot to clean up leftover files. After the uninstall the user runs the _06 install with the above command line.

In our lab environment, this scenario works perfectly every time. But as soon as it goes into production, it fails miserably and I cannot figure out what is going wrong. 1.5.0_06 will not install. I've been reading the log file line-by-line and it is possibly the strangest one I've ever seen. It begins the installation as usual, but halfway through, all of the features and components come up null.

MSI (s) (90:B4) [14:12:56:765]: Feature: other_onnonUS; Installed: Absent; Request: Null; Action: Null
And then this appears:

MSI (s) (90:B4) [14:12:56:775]: PROPERTY CHANGE: Adding REMOVE property. Its value is 'ALL'.
MSI (s) (90:B4) [14:12:56:775]: Doing action: RemoveExistingProducts
Action ended 14:12:56: InstallValidate. Return value 1.
MSI (s) (90:B4) [14:12:56:775]: Skipping RemoveExistingProducts action: current configuration is maintenance mode or an uninstall

And when it shows the command line it has extra properties I didn't pass to it - namely IS_MINOR_UPGRADE and REINSTALLMODE (server names replaced with generics).

The original Java install was packaged with InstallShield and I'm assuming there is some internal logic that is responsible for these problems. Our test lab computers use the same build as production, the same program is being run in both environments, and there is no reason why they should be behaving differently.

Any ideas? I'm at a complete loss.
0 Comments   [ - ] Hide Comments


Please log in to comment

Rating comments in this legacy AppDeploy message board thread won't reorder them,
so that the conversation will remain readable.
Answer this question or Comment on this question for clarity


Hi Megan,
this looks to me, as if there are leftovers from a previous install of v 1.5.06 present on your production system.
On the other hand, v 1.5.06 has serious bugs. Why not install v 1.5.07 ? At least as a try to see, if it's behaviour is the same.
Regards, Nick
Answered 08/07/2006 by: nheim
Tenth Degree Black Belt

Please log in to comment