we had a similar issue with 2015 and FMEServerSubmitter a while ago.
This is an extract from my logs which looks like yours:
2016-01-29 11:17:31| 1.3| 0.0|INFORM|Translation was SUCCESSFUL with 0 warning(s) (1 feature(s) output)
2016-01-29 11:17:31| 1.3| 0.0|INFORM|FME Session Duration: 0.4 seconds. (CPU: 0.2s user, 0.1s system)
2016-01-29 11:17:31| 1.3| 0.0|INFORM|END - ProcessID: 2780, peak process memory usage: 74424 kB, current process memory usage: 74424 kB
2016-01-29 11:17:31| 1.3| 0.0|INFORM|Translation was SUCCESSFUL
2016-01-29 11:17:31| 1.3| 0.0|ERROR |FMEWorkspaceRunner: The translation failed, review log for details
2016-01-29 11:17:31| 1.3| 0.0|ERROR |FMEServerJobSubmitter(ServerJobSubmissionFactory): Failed to complete run of workspace "creator2pdf2d.fmw" from repository XXXXXXXX on YYYYYYYYYYY with parameters: --SourceDataset_SHAPE """C:\ProgramData\Safe Software\FME Server\Data\zzz.shp""" --FEATURE_TYPES "" --DestDataset_PDF "C:\ProgramData\Safe Software\FME Server\Data\Boundary\bounds.pdf"
2016-01-29 11:17:31| 1.3| 0.0|ERROR |FMEServerJobSubmitter(ServerJobSubmissionFactory): Review log file "e:\apps\FMEServer\Logs\engine\current\jobs\1000\job_1346_0.log" for details
The way we fixed it was to make the child workspace generate a file with the FME status (in the shutdown python script) and test for that file on the failed port of the FMEServerJobSubmitter.
The shutdown python script code:
import fme
fout = open(FME_MacroValuesr'OUTPUT_FOLDER'] + "/Result.txt", 'w')
fout.write(str(fme.status))
fout.close()
The FMEServerJobSubmitter and extra tests:
If you open a support call with SAFE, our reference case is C111289.
Thanks Larry. I'll give it a try.
I can confirm that this is the work around to handle this situation. Thank so much for your help!