I’ve had this a few times in the last day or two, when a slave machine has restarted and stops on this error
I opened the file in question and it was not ASCII anymore, so I think something went wrong on the shutdown of the machine/slave.
Any ideas about that?
I guess I could batch replace those deadline.ini files, but as it’s just asking for a deadline repository, I kinda assumed there would be an environment variable (I’m very fond of those) that would point to the deadline repository… that would be an easier way to make sure they always know where to point
It is really odd that the file is getting corrupted, as it generally is just a really tiny text file. I can definitely pass along the suggestion of storing the repo path in an environment variable, but off the top of my head I can think of 2 or 3 customers who would have major issues with this. Can you verify if the files are all corrupted in the same fashion? If the slave was shut down properly, I can’t imagine why that would happen.
I saved one of the ini files - I’ll save another one next time it happens. I didn’t check the other ini files in the same location, not sure if they were affected.
It’s been a bit of an issue with Deadline 7 with the automatic configuration actually.
Option one is to disable autoconf (not horrible), and option two is to upgrade to the newest release of 7.1. I can’t post the download links here, but e-mail sales@thinkboxsoftware.com and they can pass it along (you’ll need a new license if you’re on 7.0).
The short of it is that there’s a bit of a juggle going on between the contents of the file and the settings, so if two Deadline apps start up quickly (Monitor and Slave for example) one may read the config while the other is re-writing the file. Nathan hit this a lot and helped us out testing it.
