OS: Windows 2008 R2
CA UIM (Nimsoft): v7.60
Initially the customer's e2e script was only running when the the configured svc_account user (domain admin account) was logged in via RDP.
Also I noticed that the RDP session/user was NOT being logged off when the probe was restarted and that indicates that the configuration is wrong.
Furthermore, the Windows Server Manager window was popping up after the user was logged on. To disable it you must enable a Windows group policy. Please refer to this link for more information.
Then RDP and login as the configured user to make sure that Server Manager window doesn't show up again.
Script is now running successfully as a local admin user on the e2e machine, e.g., ROBOT02\nimsoft01
We changed the Nimsoft Robot Watcher service to run as the local admin nimsoft01 user as well and restarted the Robot Watcher service and configured the e2e probe to run as the same user nimsoft01.
So if you use a domain admin account the Nimsoft Robot Watcher service must ALSO be running as the same user AND the e2e probe must be configured to run as this same user.
Make sure IF you change it, you must restart the Robot service after changing the 'Log On' Properties for the service (tip: run services.msc to configure the Nimsoft Robot Watcher service)
If you view the processes using the Windows Task Manager, e.g., e2e_appmon.exe and nimbus.exe (Robot), should now BOTH be running as that same admin user. ***These processes should NOT be running as 'SYSTEM.'
Also IF you use a domain admin username in the e2e probe, under the 'Run as user' setting make sure you add the DOMAIN name to the username, e.g.,
Note also that the e2e 'Run as user' field should NOT be left blank, it MUST be configured.
***NOTE that IF the probe and system are configured properly, any current RDP session(s) will be logged off when the probe is deactivated/restarted. That is a good sign that the probe is configured properly as well.
Note that you may want to 'Deactivate' the probe to kill any 'queued' script runs. Then Activate it.
Acknowledge any existing e2e alarms and open the probe and run the script using 'Start At Once.' The return code should always be 0 and the times killed should be 0 IF the script continues to run successfully. Note that you may have 'some' intermittent or random failures but it should not be more than a few out of many runs.
All issues resolved on ROBOT02 as per the description above.
We also checked a second system that was already configured to use the domain admin account not the local account, and made a few changes but both the example script and customer script did not complete and threw some e2e_starter.exe errors. This is a sign of either the domain admin account either not being configured properly (try specifying the domain account with no domain name), or some other script related issue. At that point you MUST install Real VNC server on the target machine where e2e is installed and configure a password then install VNC viewer on the machine from which you will be viewing the script runs, e.g., where you have your Infrastructure Manager installed.
Tip: Check with your Windows administrator to see if there is anything different about that account versus the local admin account in terms of system access or Group policy that may be interfering with access or script run.
keywords: ERROR e2e_starter not found will now trigger a re-logon DoLogOff running Vista? newer session handling perform logoff actions DoLogOff WTSQueryUserToken attempt reference token does not exist e2e_starter.exe