Pending (Waiting to be queued.)

[ccurwen]

Hi there

Mucking through my evaluation of the software (6.1.3), and I managed to get everything running on a pool of mini's running 10.6.7 (All are 2GHz core 2 duo's with 4G RAM)

Worker is running all of the minis, running Maya 2008, jobtype is installed and Qube! shows up in Maya menus. Supervisor is running on another mini running Mac OS X Server 10.6.7 - I installed the DB in the existing MySQL install (5.04 I think) rather than use the installer version. It all appears to be running properly:

-Supervisor comes up, shows the licenses for the eval in QubeGUI
-All of the workers show up in QubeGUI, pings

The problem is (other than it takes a while for the job to show up in QubeGUI) that nothing happens after that.  The job status just says 'pending (waiting to be queued.)'

I've tried submitting jobs from within Maya and the QubeGUI with the same result.

Can anybody tell me what I'm missing?

Thanks,

Colin

[ccurwen]

Here's an update:

Checking my supervisor log I see this error for all the jobs in my queue. This example is for the first job ID 250:

ERROR: database query error: 127.0.0.1 via TCP/IP - SELECT command denied to user 'pfx_dw'@'localhost' for table '250subjob' (1142)
SELECT id, status, host, address, data, result, count, retry, seq, macaddress, lastupdate FROM 1qube.250subjob WHERE reserve < 358029399 and status = 320 LIMIT 1

pfx_dw is the mysql user I created to access the qube DB's.  I put the username and password in the qb.conf file as detailed in the KB and I was able to start the supervisor and get the farm running:

database_user = pfx_dw
database_password = ******

Researching the MYSQL site I found that the SELECT command error can be a red herring; it throws the error because there is no table called 1qube.250subjob.

Can anyone help me out here?

[ccurwen]

selecting any job from QubeGUI on a worker machine and selecting 'execute in local shell' runs the job without issue.

still at it...!

[ccurwen]

fixed my mysql problem newbie password issue now all is well there.  Still can only re-run failed jobs directly when I execute from local shell.

I'm used to drqueue where I can trouble shoot from the frame logs - where should I be looking to discern why my workers aren't running the jobs?

[jburk]

Eval customers are entitled to use our support helpdesk system, which will give you access to a quicker turnaround on issues.

Can you send an email to support@PipelineFX.com?  This will open a ticket in our heldesk, and just reference the URL for this forum posting. Please include a contact phone number in the ticket, and we'll try to give you a call if this us something we can't sort out quickly via email.  We can also remote into your system if you d/load one of our quickSupport apps from our website under Support->QuickSupport.

You should be able to get the jobs stdout and stderr logs in the qubeGui; select a job and look in the stdout and stderr tabs in the right-hand side.