Externe Runners get stuck on every job with "Appending trace to coordinator... failed code=501"

We are running GitLab CE in an on-prem deployment. Runners within the LAN are working perfectly fine on various OSes.
Now we configured reverse proxies to make the on-prem deployment reachable from the Internet to do some magic with MacOs and Swift. Strangely though, out GitLab runners on Mac (and Linux for testing) are able to register and pick up jobs, but seem to fail to report the job status back to Gitlab.

This is what it looks like in our docker bases runner on a Mac when we try to run a sample pipeline with nothing more than a simple echo in a shell:

Using Shell executor...                             job=233 project=146 runner=********
No referees configured                              job=233 project=146 runner=********
Executing build stage                               build_stage=prepare_script job=233 project=146 runner=********
Waiting for signals...                              job=233 project=146 runner=********
Executing build stage                               build_stage=get_sources job=233 project=146 runner=********
Executing build stage                               build_stage=restore_cache job=233 project=146 runner=********
Executing build stage                               build_stage=download_artifacts job=233 project=146 runner=********
Executing build stage                               build_stage=build_script job=233 project=146 runner=********
Executing build stage                               build_stage=after_script job=233 project=146 runner=********
Executing build stage                               build_stage=archive_cache job=233 project=146 runner=********
Feeding runners to channel                          builds=1
Executing build stage                               build_stage=upload_artifacts_on_success job=233 project=146 runner=********
Skipping referees execution                         job=233 project=146 runner=********
Job succeeded                                       duration=1.4166992s job=233 project=146 runner=********
Dialing: tcp ***.*******.******.**:443 ...         
WARNING: Failed to parse "X-GitLab-Trace-Update-Interval" header  error=strconv.Atoi: parsing "": invalid syntax header-value= job=233 runner=********
WARNING: Appending trace to coordinator... failed   code=501 job=233 job-log= job-status= runner=******** sent-log=0-2011 status=501 Not Implemented update-interval=0s

Within Gitlab the job is marked as “Running” but never finished.

Does any of you have an idea how to troubleshoot this and what might be the reason for the “Appending trace to coordinator… failed” message?
From my understanding, the “Failed to parse “X-GitLab-Trace-Update-Interval” header” message is more or less cosmetic in nature and can be ignored.

Thanks in advance for your thoughts and ideas!

Cheers
Carsten