Scheduler Auto-Restart option not working

Hi,
I have been trying using the Scheduler Auto-Restart option from the "Options" menu in the "Other" tab.
The version of 24x7 we installed is for Windows 3.5.2
When setup the option, the log shows an alert message
"Restarting"
but actually it doesn't actually restart. It shuts down the timer but not the 24x7 instance.
We upgraded from version 3.4.32 where this option was running without issues.

Please, help

I guess it blocks while trying to shutdown running jobs and/or processes. Without a thorough investigations I cannot say which specific process is blocking it.

I can suggest a couple of alternative methods to force restarts and terminate processes if they don't stop.
For example, please check this solution http://www.softtreetech.com/support/phpBB2/viewtopic.php?p=20332#20332

I have tried the options mentioned in the link. It doesn't seem to work smoothly.
I have a question though. There are no jobs running in any of the servers at the time of the scheduled auto-restart. Since this feature was working in prior versions, I am wondering if this may be a bug in the newer ones.

I have tested the auto-restart in three different computers, one of them running exclusively 24x7 with only one job (a job that runs every 3 minutes, and writes a file with the date and time). I set the autorestart and there is a difference of at least 1 min. from the time the schedule job runs and the restart is scheduled. the result is the same. The clock gets disabled, but the instance doesn't shuts down and come back.
Any other ideas?

Please enable the tracing options in Tools/Options menu; Log page and let the scheduler restart (or freeze during restart). Post the contents of master.log.

I set the autorestart to 2:35 p.m.

This is the log File "schedule.log"
5/5/2009 14:28:35.281000 0 0 0 24x7 Scheduler 24x7 Scheduler starting...
5/5/2009 14:29:21.046000 0 24 0 Test File Name Job started.
5/5/2009 14:29:21.187000 0 24 0 Test File Name Job finished.
5/5/2009 14:31:21.250000 0 24 0 Test File Name Job started.
5/5/2009 14:31:21.328000 0 24 0 Test File Name Job finished.
5/5/2009 14:33:21.359000 0 24 0 Test File Name Job started.
5/5/2009 14:33:21.421000 0 24 0 Test File Name Job finished.
5/5/2009 14:35:00.875000 1 0 0 24x7 Scheduler Restarting...

There is no master.log created.

You forgot to enable the tracing options that we asked you to enabled. When enabled, 24x7 generates several additional log files, one of them is master.log and we need to see it to figure out if there is a problem with the network interface and listener.

Thanks

No, I didn't forget to enable it:
Here is a print screen of how is the configuration I set:

Please also try to enable "server" mode on the next tab (Network) settings. Restart the scheduler, let it run the test job, and then restart. Please capture screenshot of the 24x7 Trace window when the scheduler is restarting and freezing.

Also, please let us know which message is displayed in the scheduler status bar (the bottom part of the main window) when it freezes. The status bar is supposed to report on the currently performed operation.

Here are the logs created:
Master.log
-------------------
SMI10 Listener (056C4920): Start Listening: computerName(), serviceName(1096), driverName(WinSock). ... (00000000)
SMI Property: __DriverName='WinSock'
SMI Property: __ApplicationName='1096'
SMI Property: __Options='MaxListeningThreads=9'
SMI Property: MaxListeningThreads='9'
SMI AddRef SMI_TRACE(056C4348) refCount(2)
WSD11 AddRef: RefCount for listener object = 1
WSD11 AddRef: RefCount for listener object = 2
WSD11 AddRef: RefCount for listener object = 3
WSD11 AddRef: RefCount for listener object = 4
WSD11 AddRef: RefCount for listener object = 5
WSD11 AddRef: RefCount for listener object = 6
WSD11 AddRef: RefCount for listener object = 7
WSD11 AddRef: RefCount for listener object = 8
WSD11 AddRef: RefCount for listener object = 9
SMI10 Listener (056C4920): Start Listening: computerName(), serviceName(1096), driverName(WinSock). SUCCEEDED (00000000)
WSD11 AddRef: RefCount for listener object = 10
DPB30 Listener (0092FCA0): Start Listening : application(1096), location(), driver(WinSock). FAILED (80004005)
DPB55 Memory Session : (00000000) count(4331), total(638976), alloc(448356), free(172547)
DPB56 Memory Proc Common : count(1488), total(151552), alloc(82662), free(63113)
DPB31 Listener (0092FCA0): Stop Listening : ... (00000000)
SMI11 Listener (056C4920): Stop Listening: ... (00000000)
SMI11 Listener (056C4920): Stop Listening: SUCCEEDED (00000000)
WSD11 Release: RefCount for listener object = 10
WSD11 Release: RefCount for listener object = 10
WSD11 Release: RefCount for listener object = 10
WSD11 Release: RefCount for listener object = 10
WSD11 Release: RefCount for listener object = 10
WSD11 Release: RefCount for listener object = 10
WSD11 Release: RefCount for listener object = 10
WSD11 Release: RefCount for listener object = 10
WSD11 Release: RefCount for listener object = 10
WSD11 Release: RefCount for listener object = 10
SMI Release SMI_TRACE(056C4348) refCount(1)
SMI15 Listener (056C4920): Close All Connections: ... (00000000)
SMI15 Listener (056C4920): Close All Connections: SUCCEEDED (00000000)
DPB31 Listener (0092FCA0): Stop Listening : SUCCEEDED (00000000)
DPB Release DPB_LISTENER(0092FCA0) refCount(0)
SMI Release SMI_LISTENER(056C4920) refCount(0)
SMI Release SMI_TRACE(056C4348) refCount(0)
----------------------------------

The screen shot for the trace:

-----------------
The screen shot for the status bar:

Still unclear to me where it is getting stuck during shutdown. The screenshots were taken not during the restart, but before that? Am I right?

I'm trying to find out what the processing is doing when it is freezing. Would it possible to take screenshots at that time?

The screen shots were taken exactly after the instruction for restarting gets submitted. The clock on the status bar continues running, but the timer for the job executions stops. Nothing else happens. It doesn't shuts down. I can still interact with the console, but no jobs get executed. The problem is exactly that one. That you can still see the console, but the timer for running jobs stops. The instance doesn't restart and it doesn't shuts down either. The status is a kind of limbo.

I tried also installing 24x7 in an isolated machine where there is nothing else running but 24x7 with the same result. What is suspicious to me is that the old version didn't have a problem doing this. This started happening after upgrading to the 3.5.2 version

I think we are getting somewhere. Is the test job set to run detached? Right? Please flip the detached property (check or uncheck detached job property) and verify the scheduler is able to restart after that.

Same result

Are you sure? Does it restart when no jobs are running?

Is the scheduler running under low privileged user account? By the way, what type of system is this?

No it doesn't restart. The scheduler is running under an admin account. I have it running in production in a Windows Server 2003 and Windows XP.
As I mentioned before, these same systems didn't have issues with the autorestart with the version 3.4.32

Nothing else changed but the version of 24x7.