Hey,
After upgrading to the newest deadline 10.1.2.2 we are getting this error after submitting from Maya 2019.2.
Windows 10
SubmitMayaToDeadline.mel line 2737: ParseSubmissionOutputError: file …parse_output.py line 181: ParseSubmissionOutputError: Expected a blank line
SubmitMayaToDeadline.mel line 2739: Deadline: Failed to parse job ID
Maybe some one has a solution for this?
cheers,
alex
With the old submission script it´s working without any error…the problem is with the new it will only submit the first render layer and then it stops so we cannot submit all renderlayers.
cheers,
alex
Alex, did you ever find a solution to this? We’re running into the identical problem randomly
@kaiserlicht @Hybrid Was this ever figured out? We are having the same issue
Bumping this - we’re seeing this for one user at our facility but not others.
Same Maya (2019.3.1) Redshift (3.0.65) and Client (10.1.20.3) version as a user who can submit the same scene successfully.
Deadline logs are perfectly happy, but submission log has ParseSubmissionOutputErrors in line 2761 and 2763.
I know this is an old post, but I ran into that same problem with Maya 2020 and Vray 5 on Windows 10.
FIX: I ran Maya as administrator, and it worked!
Hope this helps someone,
Andrew
Our users are locked down - was running as Admin a one-off that fixed it thereafter, or does Maya need to run as admin every time for it to work?
I tested that out. It actually seems that Admin may not be necessary. It looks like (at least for me) opening Maya, then the scene from within Maya works, as opposed to opening Maya by opening the scene from its folder.
I hope that makes sense.
It makes sense, but unfortunately does not fix the problem.
OK, tracing through the code from the error messages, it’s falling over trying to parse the repository name from the DeadlineCommand output after submitting the first render layer:
# Error: file: //[our_repo]/DeadlineRepository10/submission/Maya/Main/SubmitMayaToDeadline.mel line 2761: ParseSubmissionOutputError: file //[our_repo]/DeadlineRepository10/submission/Integration/Main\deadline_submission\parse_output.py line 198: ParseSubmissionOutputError: Expected a blank line (Submitting to Repository: \\[our_repo]\DeadlineRepository10) # // Error: file: //[our_repo]/DeadlineRepository10/submission/Maya/Main/SubmitMayaToDeadline.mel line 2763: Deadline: Failed to parse job ID. //
parse_output.py looks at the first few lines of DeadlineCommand’s output to get the Deadline version and repo location. A successful submission looks like this:
_parse_version which uses _read_line to process the first line’s string to extract the deadline version, which is apparently successful.
Then it calls _parse_repository, which calls the next iteration of _read_line to get the next line of the output and errors out if it is anything but blank, returning the line it did get as part of the error message. Since we are seeing
ParseSubmissionOutputError: Expected a blank line (Submitting to Repository: \\[our repo]\DeadlineRepository10)
it looks like that blank line is missing from the output.
However, when the user manually runs deadlinecommand against the exact same submission files in his temp folder, the output looks fine:
so it seems like _read_line is getting confused and skipping a line in deadlinecommand’s output for some reason, but that doesn’t really make sense.
OK, stepping through a pared-down version of _read_line's logic in the Maya Python shell, it looks like everything’s fine, though it distinctly isn’t. At this point I’m pretty stuck.
OK - we’ve finally found a workaround/symptom. Whether the render layers job succeeds is dependent on how the Maya file was opened, which is why only 1 artist was seeing it and we couldn’t replicate even with the same file:
I think this counts as a Deadline bug, but at least there’s a workaround.
