Quote paths in docker invocation

[juliocartier]

When I try to run it through Windows on the Docker machine, it gives this error. However, I updated the python I'm running. Python is currently version 3.11.4 and still presents this error. The docker process dies.

Running cmd docker kill tpot.test.test.docker.20240526T200606.fSf0WHzreua0GuH_Z7N4ZA__
Error response from daemon: cannot kill container: tpot.test.test.docker.20240526T200606.fSf0WHzreua0GuH_Z7N4ZA__: No such container: tpot.test.test.docker.20240526T200606.fSf0WHzreua0GuH_Z7N4ZA__

Job docker.test.test.all_tasks.all_folds.TPOT failed with error: Command 'docker run --name tpot.test.test.docker.20240526T200606.fSf0WHzreua0GuH_Z7N4ZA__ --shm-size=2048M -v C:\Users\Maicon Santos.openml:/input -v C:\projetos-pythons\autobenchmark\automlbenchmark\re
sults\tpot.test.test.docker.20240526T200606:/output -v C:\Users\Maicon Santos.config\automlbenchmark:/custom --rm automlbenchmark/tpot:stable-v2.1.7 tpot test test -Xseed=auto -i /input -o /output -u /custom -s skip -Xrun_mode=docker --session=' returned non-zero exit
status 125.
Traceback (most recent call last):
File "C:\projetos-pythons\autobenchmark\automlbenchmark\amlb\job.py", line 120, in start
result = self._run()
^^^^^^^^^^^
File "C:\projetos-pythons\autobenchmark\automlbenchmark\amlb\runners\container.py", line 109, in _run
self.start_container("{framework} {benchmark} {constraint} {task_param} {folds_param} -Xseed={seed}".format(
File "C:\projetos-pythons\autobenchmark\automlbenchmark\amlb\runners\docker.py", line 76, in start_container
run_cmd(cmd, capture_error=False) # console logs are written on stderr by default: not capturing allows live display
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\projetos-pythons\autobenchmark\automlbenchmark\amlb\utils\process.py", line 281, in run_cmd
raise e
File "C:\projetos-pythons\autobenchmark\automlbenchmark\amlb\utils\process.py", line 255, in run_cmd
completed = run_subprocess(str_cmd if params.shell else full_cmd,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\projetos-pythons\autobenchmark\automlbenchmark\amlb\utils\process.py", line 98, in run_subprocess
raise subprocess.CalledProcessError(retcode, process.args, output=stdout, stderr=stderr)
subprocess.CalledProcessError: Command 'docker run --name tpot.test.test.docker.20240526T200606.fSf0WHzreua0GuH_Z7N4ZA --shm-size=2048M -v C:\Users\Maicon Santos.openml:/input -v C:\projetos-pythons\autobenchmark\automlbenchmark\results\tpot.test.test.docker.20240526
T200606:/output -v C:\Users\Maicon Santos.config\automlbenchmark:/custom --rm automlbenchmark/tpot:stable-v2.1.7 tpot test test -Xseed=auto -i /input -o /output -u /custom -s skip -Xrun_mode=docker --session=' returned non-zero exit status 125

[PGijsbers]

Looking up the error code 125 for docker run indicates that the command is invalid/contains invalid values. Based on your log, my suspicion is that the space in the mount path (between Maicon and Santos) is invalid. This can likely be resolved on our end by making sure the file paths are quoted.
On your end, you could change the benchmark configuration user_dir and input_dir in resources\config.yaml to point to paths without spaces.

If you can try a run with that configuration and can confirm that works, I would appreciate that. If my hunch is correct, I'll try to set up a patch branch for you to try out that quotes provided paths when I can find the time.

[juliocartier]

@PGijsbers I configured it in two ways and even so it still goes to the folder that has space. That's why I find it strange. I created the form following the path C:\projetos-pythons\autobenchmark\automlbenchmark and also as '{root}'. At first, when you upload the image, it recognizes the folder paths, but when you run the framework, the error continues.

Starting job docker.test.test.all_tasks.all_folds.RandomForest. [MONITORING] [docker.test.test.all_tasks.all_folds.RandomForest] Disk Usage: 71.8% Starting docker: docker run --name randomforest.test.test.docker.20240528T002142.V0ZVrW1FfJgeYEim_PCNnQ__ --shm-size=2048M -v C:\Users\Maicon Santos\.openml:/input -v C:\projetos-pythons\autobenchmark\automlbenchmark\results\randomforest.test.test.docker.20240528T002142 :/output -v C:\Users\Maicon Santos\.config\automlbenchmark:/custom --rm automlbenchmark/randomforest:stable randomforest test test -Xseed=auto -i /input -o /output -u /custom -s skip -Xrun_mode=docker --session=. Datasets are loaded by default from folder C:\Users\Maicon Santos\.openml. Generated files will be available in folder C:\projetos-pythons\autobenchmark\automlbenchmark\results. Running cmd docker run --name randomforest.test.test.docker.20240528T002142.V0ZVrW1FfJgeYEim_PCNnQ__ --shm-size=2048M -v C:\Users\Maicon Santos.openml:/input -v C:\projetos-pythons\autobenchmark\automlbenchmark\results\randomforest.test.test.docker.20240528T002142:/ou
tput -v C:\Users\Maicon Santos.config\automlbenchmark:/custom --rm automlbenchmark/randomforest:stable randomforest test test -Xseed=auto -i /input -o /output -u /custom -s skip -Xrun_mode=docker --session=docker: invalid reference format. See 'docker run --help'. Running cmddocker kill randomforest.test.test.docker.20240528T002142.V0ZVrW1FfJgeYEim_PCNnQ__`
Error response from daemon: cannot kill container: randomforest.test.test.docker.20240528T002142.V0ZVrW1FfJgeYEim_PCNnQ__: No such container: randomforest.test.test.docker.20240528T002142.V0ZVrW1FfJgeYEim_PCNnQ__

Job docker.test.test.all_tasks.all_folds.RandomForest failed with error: Command 'docker run --name randomforest.test.test.docker.20240528T002142.V0ZVrW1FfJgeYEim_PCNnQ__ --shm-size=2048M -v C:\Users\Maicon Santos.openml:/input -v C:\projetos-pythons\autobenchmark\au
tomlbenchmark\results\randomforest.test.test.docker.20240528T002142:/output -v C:\Users\Maicon Santos.config\automlbenchmark:/custom --rm automlbenchmark/randomforest:stable randomforest test test -Xseed=auto -i /input -o /output -u /custom -s skip -Xrun_mode=docker -
-session=' returned non-zero exit status 125.
Traceback (most recent call last):
File "C:\projetos-pythons\autobenchmark\automlbenchmark\amlb\job.py", line 120, in start
result = self._run()
^^^^^^^^^^^
File "C:\projetos-pythons\autobenchmark\automlbenchmark\amlb\runners\container.py", line 109, in _run
self.start_container("{framework} {benchmark} {constraint} {task_param} {folds_param} -Xseed={seed}".format(
File "C:\projetos-pythons\autobenchmark\automlbenchmark\amlb\runners\docker.py", line 76, in start_container
run_cmd(cmd, capture_error=False) # console logs are written on stderr by default: not capturing allows live display
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\projetos-pythons\autobenchmark\automlbenchmark\amlb\utils\process.py", line 281, in run_cmd
raise e
File "C:\projetos-pythons\autobenchmark\automlbenchmark\amlb\utils\process.py", line 255, in run_cmd
completed = run_subprocess(str_cmd if params.shell else full_cmd,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\projetos-pythons\autobenchmark\automlbenchmark\amlb\utils\process.py", line 98, in run_subprocess
raise subprocess.CalledProcessError(retcode, process.args, output=stdout, stderr=stderr)
subprocess.CalledProcessError: Command 'docker run --name randomforest.test.test.docker.20240528T002142.V0ZVrW1FfJgeYEim_PCNnQ --shm-size=2048M -v C:\Users\Maicon Santos.openml:/input -v C:\projetos-pythons\autobenchmark\automlbenchmark\results\randomforest.test.test
.docker.20240528T002142:/output -v C:\Users\Maicon Santos.config\automlbenchmark:/custom --rm automlbenchmark/randomforest:stable randomforest test test -Xseed=auto -i /input -o /output -u /custom -s skip -Xrun_mode=docker --session=' returned non-zero exit status 125
.
`

C:\projetos-pythons\autobenchmark\automlbenchmark\venv\Lib\site-packages\_distutils_hack\__init__.py:33: UserWarning: Setuptools is replacing distutils. warnings.warn("Setuptools is replacing distutils.") Running benchmark randomforestontestframework indockermode. Loading frameworks definitions from ['C:\\projetos-pythons\\autobenchmark\\automlbenchmark\\resources\\frameworks.yaml']. Loading benchmark constraint definitions from ['C:\\projetos-pythons\\autobenchmark\\automlbenchmark\\resources\\constraints.yaml']. Loading benchmark definitions from C:\projetos-pythons\autobenchmark\automlbenchmark\resources\benchmarks\test.yaml. Running cmddocker images -q automlbenchmark/randomforest:stableRunning cmddocker pull automlbenchmark/randomforest:stable stable: Pulling from automlbenchmark/randomforest

[juliocartier]

Is possible execute in ubuntu 22.04 in docker?

[PGijsbers]

Sorry for the slow reply, somehow this issue fell off the radar. In case it is still relevant to you or anyone else:
What seems to be happening is that the input path defaults to the openml cache directory, which defaults to a directory in your user directory, which in this case contains spaces, which again leads to the error.

Specifying the input directory explicitly to a quoted path (or one without spaces) may work. I'll try to patch the underlying issue of unquoted mounth paths.