mtor jobtype error,helps!!

[buddyred]

hi,one question,
my workstation installed RAT 6.5.2,maya 7.0
workers installed pro server and RAT,maya 7.0
qube version is 5.4
perl version is 5.8.8
also installed ActiveTcl8.5.7.0.290198
workers OS is windows server 2003 x64 edtion
when I submit a mtor job, it shows:
================================
Can't locate Class/Accessor.pm in @INC (@INC contains: C:\Program Files\pfx\qube\/../jobtypes/JobTypeLib C:\Program Files (x86)\pfx\jobtypes C:\Program Files\pfx\jobtypes C:\Program Files\pfx\qube\\types C:\Program Files\pfx\qube\api/perl/qb/blib/arch/auto/qb C:\Program Files\pfx\qube\api/perl C:/Perl/site/lib C:/Perl/lib . C:\Program Files\pfx\qube\api/perl\58) at C:\Program Files (x86)\pfx\jobtypes/mtor/GenribJob.pm line 11.
BEGIN failed--compilation aborted at C:\Program Files (x86)\pfx\jobtypes/mtor/GenribJob.pm line 11.
Compilation failed in require at C:\Program Files (x86)\pfx\jobtypes\mtor\execute.pm line 11.
BEGIN failed--compilation aborted at C:\Program Files (x86)\pfx\jobtypes\mtor\execute.pm line 11.
Compilation failed in require at C:\WINDOWS\Temp\job\0\271\271_0.pm line 11.
11.
=======================
it seems couldn't find the Class/Accessor.pm.  I update perl to 5.8.9.826.
then it shows:
=======================
mtor::GenribJob::initialize
WARNING: exception caught when trying to figure out pipe buffer size
Your vendor has not defined Fcntl macro F_SETFL, used at C:\Program Files (x86)\pfx\jobtypes/mtor/AlfredScript/TclProcessor.pm line 79.

WARNING: Assuming bufsize of [4096]
WARNING: Cannot find an executable matching exactly the version [7.0] in your PATH or in any of the system's standard locations at C:\Program Files (x86)\pfx\jobtypes/JobType.pm line 490.
WARNING: In fact, no executable file was found anywhere at C:\Program Files (x86)\pfx\jobtypes/JobType.pm line 497.
ERROR: exception caught in ->run()
ERROR: cannot find a suitable mtor executable at C:\Program Files (x86)\pfx\jobtypes/mtor/GenribJob.pm line 54.

mtor::GenribJob::initialize
WARNING: exception caught when trying to figure out pipe buffer size
Your vendor has not defined Fcntl macro F_SETFL, used at C:\Program Files (x86)\pfx\jobtypes/mtor/AlfredScript/TclProcessor.pm line 79.

WARNING: Assuming bufsize of [4096]
WARNING: Cannot find an executable matching exactly the version [7.0] in your PATH or in any of the system's standard locations at C:\Program Files (x86)\pfx\jobtypes/JobType.pm line 490.
WARNING: In fact, no executable file was found anywhere at C:\Program Files (x86)\pfx\jobtypes/JobType.pm line 497.
ERROR: exception caught in ->run()
ERROR: cannot find a suitable mtor executable at C:\Program Files (x86)\pfx\jobtypes/mtor/GenribJob.pm line 54.

ne 54.
=========================

what's the problem ?
MtoR jobtype doesn't support maya 7.0 and RAT 6.5.2??
helps.....
 

[shinya]

Hi buddyred!

For the first problem--
The Class/Accessor module should be installed with Qube 5.4, in
C:\Program Files\pfx\qube\api\perl\JobTypeLib\, when you installed
the qube core package.  Do you not see the folder and files?
If you don't see it, it could be a corrupt qube-core installation on the
worker.  Please uninstall and reinstall the qube-core software.


For the second problem, try adding the path to the installation location of
your mtor.exe program to the system PATH environment variable.
This is, for example, c:/program files/pixar/renderman-studio-1.0.1-maya2008/bin .
This will vary according to your RMS installation.

[buddyred]

Thank you shinya!
the problem has been fixed,
but new one comes.
genrib job has finished,then the render job starts
stdeer shows:
==============================================================================
----------------------------------
-     Worker 286.0
----------------------------------
requesting work for: 286.0
got work: 286:preflight - running
WARNING: exception caught when trying to figure out pipe buffer size
Your vendor has not defined Fcntl macro F_SETFL, used at C:\Program Files (x86)\pfx\jobtypes/mtor/AlfredScript/TclProcessor.pm line 79.

WARNING: Assuming bufsize of [4096]
launching: tclsh
exception caught! at C:\Program Files (x86)\pfx\jobtypes/mtor/AlfredScript/TclProcessor.pm line 137.
ERROR: error in processing alfscript: Command 'tclsh' not found in C:\Perl64\site\bin, C:\Perl64\bin, C:\Tcl\bin, C:\Python25\, C:\Perl\bin\, C:\Program Files (x86)\Alias\Maya7.0\bin, C:\WINDOWS\system32, C:\WINDOWS, C:\WINDOWS\System32\Wbem, C:\Program Files\pfx\qube\bin\, C:\Program Files\pfx\qube\lib\, C:\Program Files (x86)\Pixar\rat-6.5.2-maya7.0\bin, C:\Tcl\bin, . at C:\Program Files (x86)\pfx\jobtypes/mtor/AlfredScript/TclProcessor.pm line 134

Use of uninitialized value in subroutine entry at C:\Program Files\pfx\qube\api/perl/qb.pm line 326.
WARNING: There were at least [1] error(s)
] error(s)
] error(s)

] error(s)
===============================================================================
tcl version is ActiveTcl8.5.7.0.290198-win32-ix86-threaded,
I can't find ActiveTcl for x64
and I add "C:\Tcl\bin" to Environment Variables path.
what's the problem?

thanks again!

[shinya]

Hi buddyred,

Do you see a "tclsh.exe" program in C:\Tcl\bin\ ?

-shinya.

[buddyred]

shinya :
there only tclsh85.exe in the folder
so, is the tcl version incorrect?
or just change the file name to tclsh.exe?

[shinya]

Ah, ic, that's the issue.

You can probably rename/copy the tclsh85.exe file to tclsh.exe.

Alternatively, you can modify line 20 of the jobtypes/mtor/AlfredScript/TclProcessor.pm
file, from:

our $DEFAULT_EXECUTABLE = "tclsh";

our $DEFAULT_EXECUTABLE = "tclsh85";

and that should also do the trick.  You'll need to do that on all of your workers.

-shinya.

[buddyred]

I modified the file TclProcessor.pm on all the workers,
but two new situations:
=======================================
got work: 300:preflight - running
WARNING: exception caught when trying to figure out pipe buffer size
Your vendor has not defined Fcntl macro F_SETFL, used at C:\Program Files (x86)\pfx\jobtypes/mtor/AlfredScript/TclProcessor.pm line 79.

WARNING: Assuming bufsize of [4096]
launching: tclsh85
'$txCmd' is not recognized as an internal or external command,
operable program or batch file.
ERROR: child exited with value 1
ERROR: error in processing alfscript: ERROR: there was an error in processing nodes: ERROR: failed running system command [$txCmd -envlatl  -filter catmull-rom -sfilterwidth 3 -tfilterwidth 3 -newer "//cz3d0117x64/Patch_pillow_Proj/sourceimages/test111.tif" "//cz3d0117x64/Patch_pillow_Proj/rmantex/test111.tif.bb.tex"]


Use of uninitialized value in subroutine entry at C:\Program Files\pfx\qube\api/perl/qb.pm line 326.
WARNING: There were at least [1] error(s)
 least [1] error(s)
Error removing C:\DOCUME~1\QUBEPR~1\LOCALS~1\Temp\tclproc-Atjxs2.tcl at C:\Program Files\pfx\qube\api/perl/File/Temp.pm line 852.

le/Temp.pm line 852.
============================================
I can't find the command '$txCmd' in rattree

the second situation:
============================================
'rm' is not recognized as an internal or external command,
operable program or batch file.
ERROR: child exited with value 1
ERROR: error in processing alfscript: ERROR: there was an error in processing nodes: ERROR: failed running system command [rm "//cz3d0117x64/Patch_pillow_Proj/rib/untitled.0001.rib"]


Use of uninitialized value in subroutine entry at C:\Program Files\pfx\qube\api/perl/qb.pm line 326.
WARNING: There were at least [1] error(s)
qb.pm line 326.
WARNING: There were at least [1] error(s)
Error removing C:\DOCUME~1\QUBEPR~1\LOCALS~1\Temp\tclproc-Ybpf9a.tcl at C:\Program Files\pfx\qube\api/perl/File/Temp.pm line 852.
Error removing C:\DOCUME~1\QUBEPR~1\LOCALS~1\Temp\tclproc-cPJmFP.tcl at C:\Program Files\pfx\qube\api/perl/File/Temp.pm line 852.
l at C:\Program Files\pfx\qube\api/perl/File/Temp.pm line 852.
================================================
This situation happens in another project file.

[shinya]

Hi buddyred,

Attached is a quick hack.  You can save it and replace the AlfredScript/Node/Cmd.pm
file with it, and it should get you past the "$txCmd" and "rm" issues.  Make sure you
save it as "Cmd.pm".

So I may work on a better solution, could you post or zip and attach a ".alf" script
that gets output when you submit the job, and the "genrib" part of the job runs?
The stdout of the mtor/genrib job should print out the path to the .alf script.

-shinya.

[attachment deleted by admin]

[shinya]

buddyred,

Sorry, don't use the one from the previous post.

Use this one instead!

Thanks!
-shinya.

[attachment deleted by admin]

[buddyred]

Thank you !!
It works normally in some jobs.
I modify Cmd.pm file :
            $cmd =~ s/-p:%n/-p:1/;
====>   $cmd =~ s/-p:%n/-p:0/;
prman can use all CPU cores for one frame
 
still some ERRORs:
========================================
WARNING: Assuming bufsize of [4096]
launching: tclsh85
WARNING: skipping unknown operator [//cz3dr001/share/buttercup/rib] at C:\Program Files (x86)\pfx\jobtypes/mtor/AlfredScript/Parser.pm line 142.
script=[//cz3dr001/share/buttercup/rib]
WARNING: skipping unknown operator [//cz3dr001/share/buttercup/rib] at C:\Program Files (x86)\pfx\jobtypes/mtor/AlfredScript/Parser.pm line 183.
ERROR: error in processing alfscript: ERROR: there was an error in processing nodes: Can't locate object method "process" via package "mtor::AlfredScript::Node:://cz3dr001/share/buttercup/rib" at C:\Program Files (x86)\pfx\jobtypes/mtor/AlfredScript.pm line 90.


Use of uninitialized value in subroutine entry at C:\Program Files\pfx\qube\api/perl/qb.pm line 326.
WARNING: There were at least [1] error(s)
st [1] error(s)
Error removing C:\DOCUME~1\QUBEPR~1\LOCALS~1\Temp\tclproc-71LiV9.tcl at C:\Program Files\pfx\qube\api/perl/File/Temp.pm line 852.
Error removing C:\DOCUME~1\QUBEPR~1\LOCALS~1\Temp\tclproc-VBz5fZ.tcl at C:\Program Files\pfx\qube\api/perl/File/Temp.pm line 852.
.
i/perl/File/Temp.pm line 852.
==============================================
what's the problem?
=====================
when the render job starts, It creat a lot of "perl.exe" "tcl85.exe" processes on the workers and increasing with the number of frames.
I submit a job with 20 frames, it starts rendering, there's 4 "perl.exe" processes on the worker, To the 15th frame , 50 "perl.exe" processes...and many "tcl85.exe"
and seems to more CPU cores more the processes.
   
any solutions?

[shinya]

----
script=[//cz3dr001/share/buttercup/rib]
WARNING: skipping unknown operator [//cz3dr001/share/buttercup/rib] at C:\Program Files (x86)\pfx\jobtypes/mtor/AlfredScript/Parser.pm line 183.
ERROR: error in processing alfscript: ERROR: there was an error in processing nodes: Can't locate object method "process" via package "mtor::AlfredScript::Node:://cz3dr001/share/buttercup/rib" at C:\Program Files (x86)\pfx\jobtypes/mtor/AlfredScript.pm line 90.
----

Could you find and post the .alf script that's generating that error?
Again, you should be able to find the script's path in the stdout.


As for your perl.exe question-- it's normal to see at least one perl.exe instance
per subjob.  If your workers have the default setting, they're set up to run
as many subjobs as there are cores.  So, for example, if your worker has 4 cores,
then it can run up to 4 subjobs simultaneously. 

If you're running a multithreaded app, such as renderman, you may want only
one subjob to run at any given time.  To do that, when you submit your job,
specify a "reservation" of "host.processors=1+" (yes, literally "1+") so that each
subjob will occupy all of the cores on the workers. 

Alternatively, you can set up your workers to only take one subjob at a time.
You can do so in the Configuration GUI (Start > All Programs > PipelineFX > PipelineFX Qube! 5.4 > Configuration),
in the "Worker Settings" tab > Advanced Settings > Cpus (Override Detected).  Set
that to 1.


Anyway, please post the .alf script in question, so I may reproduce the issue.
Thanks!

[buddyred]

here is the .alf script in question:
===============================
//cz3dr001/share/buttercup/rib
===============================


this is normally .alf script
==================================
Task -title {Frame.0002} -subtasks {
   Task -title {FramePreflight.0002} -subtasks {
      Task -title {Textures.0002} -subtasks {
         } -cmds {
         } -cleanup {
         } -chaser {
         } -preview {
         }
      Instance {JobPreflight}
      } -cmds {
      } -cleanup {
      } -chaser {
      } -preview {
      }
   } -cmds {
      Cmd {prman -p:%n -Progress  \"//cz3d0117x64/Patch_pillow_Proj/rib/zd_sc_005_fangzi_bg.0002.rib\"} -service {local:pixarRender} -tags {prman intensive}
   } -cleanup {
   } -chaser {
   } -preview {
      sho \"//cz3d0117x64/Patch_pillow_Proj/rmanpix/zd_sc_005_fangzi_bg.0002.iff\"
   }
==============================================



I already set cpus (Override Detected) to 1, I know this is useful for maya software render and mentalray render, but useless for prman.

still a lot of "perl.exe"
here is the screenshot perl.jpg




[attachment deleted by admin]

[shinya]

Hi there,

I don't have any issues parsing the .alf script you posted.
Can you send me the file itself as an attachment (not copy/paste)?

Also, could you do me a favor, and send a zipped joblog of the mtor jobs
(the genrib part and the render part) that failed? 

The instructions on finding and sending them is included below.
Please send them to support <at> pipelinefx.com .

btw, the number of perl.exe running on your system does look bogus.
They're probably stale processes from jobs that didn't quite clean up
properly.  Make sure you don't have jobs running on the worker, and
kill those processes (or reboot the system).  Then, if you can find out
a pattern in which those perl.exe processes build up (such as what jobtypes
cause them), that would help us determining the cause and fixing the issue.

Thanks!

----
In order to address your problem more completely, please send us the job log directory for the job in question.

You can locate the directory by logging into your Supervisor, and looking for the job log folder in the following location (depending upon your Supervisor platform):

Windows XP
\Program Files\pfx\qube\logs\job

Windows 2008/Vista
\ProgramData\Pfx\Qube\logs\job

Linux, OS X
/var/spool/qube/job

In that folder, you will find a numbered directory that corresponds to the number of thousands in the job ID. (ID < 1000 = 0) Search in one of these folders for the one that corresponds to the correct job ID.

Zip up the entire job ID folder and reply to this email message with the zipped file.

If the zipped job ID folder turns out to be larger than 2MB, don't send it, but let us know and we will help you address the problem in an alternative fashion.