I'm not familiar with the lotus notes silent install switches, but if they are incorrect that's the most logical explanation. If you run the same exe with the same switches in a command prompt, is it silent or do you see the UI?
I'm not following what your saying. Are you trying to get me to create a command promt package with the same switches> If so, I can't see the setup.exe file that I need to push?
No, I'm suggesting that you test the package by running it manually in command prompt. Completely outside of Ivanti. Run it manually using the same switches and make sure it installs successfully. If it won't install manually, it won't install via Ivanti either. This is the best practice when developing any new deployment package.
If it installs successfully when executed manually in cmd with the same switches, then make sure your ivanti deployment package is configured to use the same account and try it again. You could also post the c:\program files (x86)\landesk\ldclient\data\sdclient_taskxxx.log (replace xxx with the task ID number for your specific task) and I'd be happy to review it.
1 of 1 people found this helpful
Hello these are the switches I used to deploy the Notes fix pack 6 /S /v/qn
Im trying to install Fix Pack 7 - I tried to use the silent install /S /v/qn and setup.exe /S /v/qn But still I got an error return code of 1603
I investigated the log you posted, and it's showing that it successfully installed the program. The last lines in the log indicate as such:
Fri, 11 Aug 2017 09:26:12 Execute Msiexec.exe with command Line: "Msiexec.exe" /quiet /norestart /i "C:\Program Files (x86)\LANDesk\LDClient\sdmcache\SWD\Lotus Notes\9.0.1\Multi User Install\IBM Notes\IBM Notes 9.0.1 Social Edition.msi" REBOOT=ReallySuppress
Fri, 11 Aug 2017 09:32:40 MSI result DB50000
Fri, 11 Aug 2017 09:32:40 Rebooting system if needed
Fri, 11 Aug 2017 09:32:44 processing of package is complete, result 229965824 (0x0db50000 - code 0)
For an MSI return code DB50000 maps to "0" which is a success return code. Did the program actually install? Or, perhaps was the program already installed on that machine? Maybe uninstalled, but it left behind some reg keys? .MSI installers will detect if the package is already installed and if so will usually return a success code without actually installing anew.
I have gathered new log files: sdclient_task1860.log - Google Drive
Mon, 14 Aug 2017 08:47:15 ExpandEnvironmentVariables Result: /S /v/qn
Mon, 14 Aug 2017 08:47:15 LSWD or Executable Client Thread
Mon, 14 Aug 2017 08:47:15 PackagePath: [http://UK01B144.uk01.veka.com/SWD/Lotus Notes/9.0.1 FixPack/FP7/setup.exe]
Mon, 14 Aug 2017 08:47:15 Processing generic executable
Mon, 14 Aug 2017 08:47:23 Launched application 'C:\Program Files (x86)\LANDesk\LDClient\sdmcache\SWD\Lotus Notes\9.0.1 FixPack\FP7\setup.exe' ('/S /v/qn') result 1603
Mon, 14 Aug 2017 08:47:23 Installation result 8DB50643
Mon, 14 Aug 2017 08:47:23 RunPackageInstall: stop on returncode=8db50643 of package=Install_Notes_FP_v7
Mon, 14 Aug 2017 08:47:23 processing of package is complete, result -1917516221 (0x8db50643 - code 1603)
Now, this package installs on some computers but now all computers.
Check whether the package is an MSI wrapped in an exe, or whether it's an "actual EXE" ... and turn on verbose logging for the installer.
The exit code (1603) and the switches strike me as very MSI-ish - thus my suspicion. We're "only the messenger" in this particular case - so if installs are failing, your first port of call should always be the application/installer log (which I hope has some kind of options around verbosity / debug-logging).
If we're talking MSI that's wrapped in an exe, here's a few helpful places for 1603 codes:
<The last link is fine, even though it looks a bit cut-off...>
Thank you for your fast response. I kinda new this morning that Ivanti/ LanDESK was correct and it was pushing the package out.
It was failing because the Notes client its self was open on the end machine. In fact, the client should of closed the Notes application for the FP to be installed. I've got it successful now.
With stuff like this, you *CAN* push out the package via a custom vulnerability.
The benefit here is that you can (aggressively) kill off running processes before applying the patch (it's a think that's commonly needed for Java & Flash for instance).
Merely pointing out that the possibility exists -- obviously end-users should be forewarned before you close stuff on them harshly .
How can I complete this?