Hi Synapse Team,
My submitted docker image fails to run, as it gives no *log.txt in the log-files.
It creates only the *log.zip file, which is pretty hard for me to troubleshoot.
On the local machine, the docker container runs smoothly with:
```
docker run -v ":/input/" -v ":/output/" docker.synapse.org/syn21746389/sc1_beataml:version4
```
The error message is:
```
Your workflow job, (submission ID 9701879), has failed to complete. The message is:
-packages/toil/leader.py", line 246, in run
STDERR: 2020-03-12T11:02:16.960627394Z raise FailedJobsException(self.config.jobStore, self.toilState.totalFailedJobs, self.jobStore)
STDERR: 2020-03-12T11:02:16.960638907Z toil.leader.FailedJobsException
```
Thanks for your help !
Created by Alexander Ohnmacht aljoshoh @Nova,
I'll have to look into your submissions a bit more. I'll get back to you as soon as I can!
Best,
Verena @aljoshoh,
> ...where I aimed at outputting at least the *log.txt file
Docker run log files are not available when submitting to the Validation queues, as noted in the announcement [here](https://www.synapse.org/#!Synapse:syn20940518/discussion/threadId=6803). You can, however, use the Leaderboard queue to test your container! These queues will continue to return the log files.
> Are you mounting a directory on /home when running it on the synapse server ? I am asking because I have my models and the entrypoint script located at /home in the container.
Directories are mounted into the root directory, but I don't think that should affect your entrypoint script! So long as your script is calling for the input files using the absolute path and that the output file is created inside `/output`.
Hope this helps!
Verena @Nova @v.chung,
I just double-checked, I also can run successfully on the leaderboard , with respective *log.txt !
Why could this be ? Hi Verena @v.chung ,
I have a similar problem when I submit the model to validation phase. The same model works when submitted to the leaderboard. Here is the error:
```
Your workflow job, (submission ID 9701977), has failed to complete. The message is: -packages/toil/leader.py", line 246, in run STDERR: 2020-03-13T05:09:42.036112591Z raise FailedJobsException(self.config.jobStore,
self.toilState.totalFailedJobs, self.jobStore) STDERR: 2020-03-13T05:09:42.036125457Z toil.leader.FailedJobsException
```
Thank you so much for your help!
@v.chung The example above is a minimal example, where I aimed at outputting at least the *log.txt file
In my most recent submission (which should create predictions.csv), I get the same error with a similar *log.zip:
```
Your workflow job, (submission ID 9701879), has failed to complete. The message is:
-packages/toil/leader.py", line 246, in run
STDERR: 2020-03-12T11:02:16.960627394Z raise FailedJobsException(self.config.jobStore, self.toilState.totalFailedJobs, self.jobStore)
STDERR: 2020-03-12T11:02:16.960638907Z toil.leader.FailedJobsException
```
Running the image locally, it creates predictions.csv successfully. Are you mounting a directory on /home when running it on the synapse server ? I am asking because I have my models and the entrypoint script located at /home in the container. I am not sure how to troubleshoot this, since the docker container does not seem to even start running.
Thank you for your efforts ! Dear @aljoshoh,
You're not alone - the zipped log file definitely takes some getting used to! I looked at your submission logs; looks like the expected output file, `predictions.csv`, is not being created:
```
Exception: No 'predictions.csv' file written to /output, please check inference docker
```
When you locally run your container, do you get that output file?
Let me know,
Verena