Error when deploying example repo in the real robot

Hello,

I am trying to learn how to submit code to the real robot. I am using for now the rrc_example_package.

After successfully testing it locally following the instructions in
How to Locally Run Your Code in Simulation — Real Robot Challenge 2021 documentation” I am trying to submit it to the real robot server. I follow the instructions in “Job Submission System — Real Robot Challenge 2021 documentation”.

my roboch.json file is validated correctly when I run the “check” on the realrobot server. I have just forked and submitted this “rrc_example_package” repo, but the execution fails with the following error:

Submission failed with the following error:
Command ‘[‘singularity’, ‘exec’, ‘–cleanenv’, ‘–contain’, ‘–net’, ‘–network’, ‘none’, ‘-B’, ‘/var/lib/condor/execute/dir_7095/run_submission-c3vf7p6r:/ws’, ‘/beegfs/pool2/nethome/solemnlollies/payload/’, ‘bash’, ‘-c’, ‘. /setup.bash; cd /ws; colcon build’]’ returned non-zero exit status 255.

I thought the repo worked out of the box, and indeed it does work when I test it locally. Maybe am I missing something?

Thank you very much for any help

Hello,

I could not replicate the error. I uploaded this roboch.json file:

{
  "repository": "https://github.com/rr-learning/rrc_example_package.git",
  "email": "myname@myaffiliation.de"
}

and after ssh-ing to the submission system, I called “submit”. A bit later I received a mail with the data, and no report of error.

Would you have some hints on what you may have done differently, so I can try it out ?

Vincent

I can replicate your results successfully! my error was that my roboch.json file contained the field

“singularity_image”: “”,

And the command “check” in the robot server was telling me that it was fine. Thank you very much for you help.

Hey,

I get a similar error:

Command '['singularity', 'exec', '--cleanenv', '--contain', '--net', '--network', 'none', '-B', '/var/lib/condor/execute/dir_14477/run_submission-h_5c6_yr:/ws', '/beegfs/pool2/nethome/username/payload/user_image.sif', 'bash', '-c', '. /setup.bash; cd /ws; colcon build']' returned non-zero exit status 1.

All checks were passed

email address passed ✔
custom singularity image found passed ✔
custom singularity image integrity passed ✔
git repository URL passed ✔
git deploy key found passed ✔
access repository with deploy key passed ✔

We are using a custom image
-rwxr-xr-x 1 username 2.7G Thu 07/08/2021 17:54 user_image.sif*

Locally everything runs fine.

Any idea what could be wrong?

Best,
Sebastian

@sblaes

I fail to replicate the issue. I submitted a job with the default singularity image, and things worked as expected.

Here the content of my roboch.json file:

{
    "repository": "https://github.com/rr-learning/rrc_example_package.git",
    "email": "myname@myaffiliation.de",
    "singularity_image": "rrc2021_latest.sif"
}

May I ask you to also try submitting with the default image ?

https://people.tuebingen.mpg.de/felixwidmaier/rrc2021/singularity.html#singularity-download-image

@vincent.berenz

Thanks for looking into it.

Yes, the default image works.

But I can’t run my code without the custom image. Would it be possible to provide some more detailed error message from my job? I’m not sure at which point the job fails.

Best,
Sebastian

Thanks for testing with the basic image, that helps excluding some other issues. Indeed, more info would be needed, we will look into this