Examples: query, "exact match", wildcard*, wild?ard, wild*rd
Fuzzy search: cake~ (finds cakes, bake)
Term boost: "red velvet"^4, chocolate^2
Field grouping: tags:(+work -"fun-stuff")
Escaping: Escape characters +-&|!(){}[]^"~*?:\ with \, e.g. \+
Range search: properties.timestamp:[1587729413488 TO *] (inclusive), properties.title:{A TO Z}(excluding A and Z)
Combinations: chocolate AND vanilla, chocolate OR vanilla, (chocolate OR vanilla) NOT "vanilla pudding"
Field search: properties.title:"The Title" AND text
Answered
How Come

How come PipelineController has changed in a very backwards-incompatible fashion but we are still on the same major version? O_O

Also the documentation is not reflective of this

  
  
Posted 3 years ago
Votes Newest

Answers 30


Does it mean that if it is set to False I need an agent but if I set it to True I don't need one?

  
  
Posted 3 years ago

That was the idea behind the feature (and BTW any feedback on usability and debugging will be appreciated here, pipelines are notorious to debug 🙂 )

the ability to exexute without an agent i was just talking about thia functionality the other day in the community channel

What would be the use case ? (actually the infrastructure now supports it)

  
  
Posted 3 years ago

awesome martin, thanks

  
  
Posted 3 years ago

Sorry but its not clear

  
  
Posted 3 years ago

Yes very much

  
  
Posted 3 years ago

Is this a common case? maybe we should change the run_pipeline_steps_locally argument to False?
(The idea of run_pipeline_steps_locally=True is that it will be easier to debug the entire pipeline on the same machine)

  
  
Posted 3 years ago

Not sure I understand, if i run pipe.start_locally(run_pipeline_steps_locally=True|False) what is the difference betwee ntrue and false? assuming I want to execute locally

  
  
Posted 3 years ago

the ability to exexute without an agent i was just talking about thia functionality the other day in the community channel

  
  
Posted 3 years ago

awesome this will help a lot with debugging

  
  
Posted 3 years ago

Ohh, sorry 🙂
:param run_pipeline_steps_locally: (default False) If True, run the pipeline steps themselves locally as a subprocess (use for debugging the pipeline locally, notice the pipeline code is expected to be available on the local machine)

  
  
Posted 3 years ago

ShinyLobster84

  
  
Posted 3 years ago

Hi Martin,
I upgraded the ClearML version to 1.1.1 and I updated the pipeline code according to v2 as you wrote here and I got a new error which I haven't got before.
Just noting that I did a git push before.
Do you know what can cause this error?
Thanks!
version_num = 1c4beae41a70c526d0efd064e65afabbc689c429 tag = docker_cmd = ubuntu:18.04 entry_point = tasks/pipelines/monthly_predictions.py working_dir = . Warning: could not locate requested Python version 3.8, reverting to version 3.6 created virtual environment CPython3.6.9.final.0-64 in 648ms creator CPython3Posix(dest=/root/.clearml/venvs-builds/3.6, clear=False, no_vcs_ignore=False, global=True) seeder FromAppData(download=False, pip=bundle, setuptools=bundle, wheel=bundle, via=copy, app_data_dir=/root/.local/share/virtualenv) added seed packages: pip==21.2.4, setuptools==58.1.0, wheel==0.37.0 activators BashActivator,CShellActivator,FishActivator,NushellActivator,PowerShellActivator,PythonActivator cloning: fatal: could not read Username for ' ': terminal prompts disabled Repository cloning failed: Command '['clone', ' ` ', '/root/.clearml/vcs-cache/my_project.git.64cf56a47b594c89c550b96afe67e25f/my_project.git', '--quiet', '--recursive']' returned non-zero exit status 128.
clearml_agent: ERROR: Failed cloning repository.

  1. Make sure you pushed the requested commit:
    (repository=' ', branch='monthly_pipeline', commit_id='1c4beae41a70c526d0efd064e65afabbc689c429', tag='', docker_cmd='ubuntu:18.04', entry_point='tasks/pipelines/monthly_predictions.py', working_dir='.')
  2. Check if remote-worker has valid credentials [see worker configuration file]
    2021-10-05 10:33:39
    Process failed, exit code 1 `
  
  
Posted 3 years ago

ShinyLobster84

fatal: could not read Username for '

': terminal prompts disabled

This is the main issue, it needs git credentials to clone the repo code, containing the pipeline logic (this is the exact same behaviour as pipeline v1 execute_remotely(), which is now the default, could it be that before you executed the pipeline logic, locally ?)
WackyRabbit7 could the local/remote pipeline logic could apply in your case as well ?

  
  
Posted 3 years ago

we are running the agent on the same machine AgitatedDove14 , it worked before upgrading the clearml... we never set these credentials

  
  
Posted 3 years ago

and we want to continue xecuting locally

  
  
Posted 3 years ago

I understand it starts locally... what does it mean to run the steps locally?

  
  
Posted 3 years ago

Worth mentioning, nothing has changed before we executed this, it worked before and now after the update it breaks

  
  
Posted 3 years ago

We try to break up every thing into independent tasks and group them using a pipeline. The dependency on an agnet caused an unnecessary overhead since we just want to execute locally. It became a burden once new data scientists join the project and instead of just telling them "yeah, just execute this script" you have to now teach them about clearml, the role of agents, how to launch them, how they behave, how to remove them and stuff like that... things you want to avoid with data scientists

  
  
Posted 3 years ago

WackyRabbit7 interesting! Are those "local" pipelines all part of the same code repository? do they need their own environment ?
What would be the easiest pipeline interface to run them locally? (I would if we could support this workflow, it seems you are not alone in this approach, and of course that you can always use them remotely, i.e. clone the pipeline and launch it on an agent)

  
  
Posted 3 years ago

And I'm getting this in the command line when the process fails:
/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.8/lib/python3.8/multiprocessing/resource_tracker.py:216: UserWarning: resource_tracker: There appear to be 73 leaked semaphore objects to clean up at shutdown warnings.warn('resource_tracker: There appear to be %d '

  
  
Posted 3 years ago

AgitatedDove14

  
  
Posted 3 years ago

WackyRabbit7

we did execute locally

Sure, instead of pipe.start() use pipe.start_locally(run_pipeline_steps_locally=False) , this is it 🙂

  
  
Posted 3 years ago

I've read that

  
  
Posted 3 years ago

I started running it again and it seems to have passed the phase where it failed last time

Yey!

Yes it is a common case....

I have the feeling ShinyLobster84 WackyRabbit7 you are not alone in this one 🙂 let me make sure we change the default value of Yes it is a common case to False, so the code looks cleaner

  
  
Posted 3 years ago

Yes it is a common case but I think the pipe.start_locally(run_pipeline_steps_locally=False) solved it. I started running it again and it seems to have passed the phase where it failed last time

  
  
Posted 3 years ago

But what does it do?

  
  
Posted 3 years ago

what is the difference betwee nfalse and true?

  
  
Posted 3 years ago

what does it mean to run the steps locally?

start_locally : means the pipeline code itself (the logic that runs / controls the DAG) runs on the local machine (i.e. no agent), but this control logic creates/clones Tasks and enqueues them, for those Tasks you need an agent to execute them
run_pipeline_steps_locally=True: means the Tasks the pipeline creates, instead of enqueuing them and having an agent runs them, they will be launched on the same local machine (think debugging, otherwise I cannot actually see the value). And to your question yes it means there is no need for an agent.
WackyRabbit7 did that help?

  
  
Posted 3 years ago

we did execute locally

  
  
Posted 3 years ago

if I called .start_locally

  
  
Posted 3 years ago
975 Views
30 Answers
3 years ago
one year ago
Tags