Scheduler not running after migration to Linux

[dgshue]

OTRS 3.3.7
CentOS 6.5
MS-SQL 2008 R2
This is perl, v5.10.1 (*) built for x86_64-linux-thread-multi

Just got through with a migration for Windows 2008 R2/Apache/Strawberry Perl to CentOS. Copied my ZZAAuto.pm and Config.pm and that's all. Did a clean 'yum install' of OTRS on the CentOS box. Upgraded to 3.1 -> 3.2 -> 3.3.7 cleanly and running all applicable scripts in the process. OTRS seems to be running fine.

However, the GUI is complaining that the Scheduler is not running. I have tried firing it up from the GUI and the CLI as root and otrs users and it does not seem to be updating the DB process_id table. DashboardStats and Postmaster are running and seem to be working fine. The process it self seems to be starting up, but the GUI still reports the error and -a status reports "Not Running!". I assume this is due to the fact that the PID is not listed in the DB. I've enabled Debug=5, but none of the logs really give any useful information about the Scheduler process. Does anyone have any suggestions as to where I can look and/or some insight into the Scheduler process?

root 6597 1 0 09:13 ? 00:00:00 /usr/bin/perl /opt/otrs/bin/otrs.Scheduler.pl -a start
root 6601 1 0 09:13 ? 00:00:00 /usr/bin/perl /opt/otrs/bin/otrs.Scheduler.pl -a start
otrs 6611 1 0 09:15 ? 00:00:00 /usr/bin/perl /opt/otrs/bin/otrs.Scheduler.pl -w 1
root 6620 1 0 09:17 ? 00:00:00 /usr/bin/perl /opt/otrs/bin/otrs.Scheduler.pl -a start
otrs 6641 1 0 09:20 ? 00:00:00 /usr/bin/perl /opt/otrs/bin/otrs.Scheduler.pl -w 1
root 6647 1 0 09:21 ? 00:00:00 /usr/bin/perl /opt/otrs/bin/otrs.Scheduler.pl -a start
otrs 6663 1 0 09:25 ? 00:00:00 /usr/bin/perl /opt/otrs/bin/otrs.Scheduler.pl -w 1
otrs 6686 1 0 09:25 ? 00:00:00 /usr/bin/perl /opt/otrs/bin/otrs.Scheduler.pl -a start

[jojo]

any errors on the database level. Please note that MS SQL with Linux as application server is not supported.

[dgshue]

Running profiler while I start the Scheduler shows this:

SET DATEFORMAT ymd

set fmtonly on select process_name from process_id where 1=2 set fmtonly off

declare @p1 int
set @p1=1
exec sp_prepexec @p1 output,N'@P1 nvarchar(14)',N'
SELECT process_name, process_id, process_host, process_create, process_change
FROM process_id
WHERE process_name = @P1',N'otrs.Scheduler'
select @p1


I never see an INSERT/UPDATE statement, it appears to only be a status check. Anyway to debug the script?

[dgshue]

Ok, new information on this case. After editing the otrs.Scheduler.pl to not damonize, it seems to be running fine. Updates the DB and so the GUI is not complaining. Must be having a problem with the Daemon.pm... tried running through a 'perl -d', when it forks I loose the trace and I'm lost. Anyone seen anything like this?

[dgshue]

More information, while running the Scheduler "interactively" if you will... it will run smoothly for 5 minutes until the new 'scheduler-watchdog' fires off from cron and then I receive this error:

ERROR: OTRS-otrs.Scheduler-10 Perl: 5.10.1 OS: linux Time: Fri Jun 20 14:10:02 2014

Message: Process could not be found in the process table!
Scheduler is stopping...!


Traceback (2680):
Module: main::_AutoStop (unknown version) Line: 672
Module: main::_Start (unknown version) Line: 443
Module: ./otrs.Scheduler-Test.pl (unknown version) Line: 161

I've also noticed that overtime, the watchdog process will also 'hang' and fire off multiple processes and consume most of my memory before the night is out. This morning, OTRS didn't respond at all as Apache could not consume any more memory due the amount of 'watchdog' processes.

otrs 2707 1 0 14:10 ? 00:00:00 /usr/bin/perl -w /opt/otrs/bin/otrs.Scheduler.pl -w 1
otrs 2744 1 0 14:15 ? 00:00:00 /usr/bin/perl -w /opt/otrs/bin/otrs.Scheduler.pl -w 1
otrs 2773 1 0 14:20 ? 00:00:00 /usr/bin/perl -w /opt/otrs/bin/otrs.Scheduler.pl -w 1
otrs 2797 1 0 14:25 ? 00:00:00 /usr/bin/perl -w /opt/otrs/bin/otrs.Scheduler.pl -w 1

[dgshue]

Bump

[catweazle]

Code: Select all

Upgraded to 3.1 -> 3.2 -> 3.3.7

1. why?!

2. what does /var/log/messages tell?

[dgshue]

/var/log/messages -> nothing reported pertaining to Scheduler
/opt/otrs/var/log/otrs.log -> nothing about Scheduler when running as Daemon / running interactively does show PID (runs normally)
/opt/otrs/var/log/SchedulerOUT.log -> is created but empty

[dgshue]

Code: Select all

Upgraded to 3.1 -> 3.2 -> 3.3.7

1. why?!

Per 'UPGRADING' documentation

Code: Select all

If you are running a lower version of OTRS you have to follow the upgrade path
to 3.2 first (1.1->1.2->1.3->2.0->2.1->2.2->2.3->2.4->3.0->3.1->3.2->3.3 ...)!
You need to perform a full upgrade to every version in between, including
database changes and the upgrading perl script.

[Giulio Soleni]

Hi,
as per viewtopic.php?f=62&t=17742 it might be an issue due to multiple scheduler process records stuck in the database.
Try to stop all OTRS processes and clean up process_id table.

HTH

[dgshue]

I hate to keep saying "I tried that", but I tried that. Running as a daemon, it just doesn't seem to do anything.