ERROR:Attempted to access non-existent config key PrepForOFX

Deadline
1

Hey Guys,

All of a sudden this AM we are getting this error on any Nuke comp we submit. We have not had any issues and literally the same files we submitted yesterday which rendered fine are now kicking out the error below. The only thing that happened last night is it looks like we have have had a power fluctuation but everything else is working fine. This is Deadline Version: 6.1.0.54665 R on Win7 clients and Linux Repository

Thanks,
Andrew

=======================================================
Error

Error in StartJob: GetBooleanConfigEntry: Attempted to access non-existent config key PrepForOFX (Deadline.Plugins.RenderPluginException)
at Deadline.Plugins.ScriptPlugin.GetBooleanConfigEntry(String key)
at Deadline.Plugins.DeadlinePlugin.GetBooleanConfigEntry(String key)
at Python.Runtime.Dispatcher.TrueDispatch(ArrayList args)
at Python.Runtime.Dispatcher.Dispatch(ArrayList args)
at Deadline.Plugins.DeadlinePlugin.StartJob()
at Deadline.Plugins.ScriptPlugin.StartJob(Job job, String& outMessage, AbortLevel& abortLevel)
at Deadline.Plugins.ScriptPlugin.StartJob(Job job, String& outMessage, AbortLevel& abortLevel)

=======================================================
Type

RenderPluginException

=======================================================
Stack Trace

at Deadline.Plugins.Plugin.StartJob(Job job)
at Deadline.Slaves.SlaveRenderThread.RenderCurrentTask(TaskLogWriter tlw)

2

I turned off the OFX Cache settings in the Plugin Configuration and things seem to be going now, but not sure why I would have to change that all of a sudden, and I don’t know the ramifications of turning that off.

3

Hi Andrew,
This sounds very much like your “Nuke.dlinit” file became corrupted during your recent power outage, as the config key “PrepForOFX” should be present in the Nuke.dlinit file. By you setting it to False in the Plugin Configuration, it effectively was re-created again by the UI writing the key back out to the Nuke.dlinit file. If I’m right, you could now change this back to “True” again and you should be back to where you were before. By default we do ship with this setting set to True. I will also make a tweak to our Nuke plugin to handle this edge case, so it doesn’t throw any plugin error if this situation occurs for anyone else in the future.

My only concern, is that you may have other corruptions throughout your setup as a result of this power outage? Did the Deadline repository get shutdown in an un-graceful manner? Do you have a backup of the repository that you could use to selectively restore some parts from?

Mike

4

Hi Mike,

I found out the power outage definitely made some issues in other places. We lost all of our settings in the plug-ins configurations so we were pointing to the wrong versions of software. I’ve gone through pretty much all of the other settings and everything seems ok so far. Unfortunately I don’t have a backup of the repository, but it sounds like I should probably be making daily backups.

Thanks…
Andrew

5

One thing I’m noticing now is I can’t get the slaves to pick up more than one concurrent task. I changed the slave setting to 5 as well as have the Limit to Slave settings turned off and I"m still just getting one instance. Ideas?

6

I have a feeling you may have a database corruption here as well. Is your MongoDB running on the same machine that the Deadline Repository is on, which suffered this power outage?

Do you have a backup of your MongoDB?

There are a couple of options to resolve the DB issue if it turns out to be the cause of the issue. Can you send your DB logs found in the Mongo directory (zipped) up to our support ticket system and the guys will be able to advise on a database repair command that can be run if necessary.

7

I don’t have a backup of the MongoDB either… was running on the same system.

I’m sending the logs now…

Thanks.