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
Hi Everyone, I’D Really Appreciate Your Advice—Thank You In Advance! I’Ve Created A Pipeline Consisting Of Tasks. I Ran Tasks And They Are In Status “Draft”. When I Run The Pipeline Script Using Pipe.Start_Locally(), Everything Works Fine, And The Pipelin

Hi everyone, I’d really appreciate your advice—thank you in advance! I’ve created a pipeline consisting of tasks. I ran tasks and they are in status “Draft”. When I run the pipeline script using pipe.start_locally(), everything works fine, and the pipeline completes successfully. However, when I use pipe.start(queue='default'), the pipeline remains in the "running" status, but the cloned task stays in the "queued" status and never executes. What could be the issue, and how can I resolve it? 🥲

  
  
Posted 2 months ago
Votes Newest

Answers 13


#run this cmd will get you one more default queue
clearml-agent daemon --queue default --detached
  
  
Posted 2 months ago

clearml-agent daemon --queue default --stop

  
  
Posted 2 months ago

FrustratingGoldfish63 thank you VERY much!!! It works ♥ 🫶 🔥

  
  
Posted 2 months ago

Wow! Ok! I’ll try this ♥ 🫶 thank you very much for your advices!

  
  
Posted 2 months ago

Thank you for the answer ! So, should I run this command twice, but with different names?

  
  
Posted 2 months ago

one machine can attach multi queues
I use default queue for pipeline and services queue for task but two default also work

  
  
Posted 2 months ago

Thank you so much ! ☺

  
  
Posted 2 months ago

Oh, thank you very much! ♥ So, should we always have two queries for the pipeline and tasks?
P.S. I try to run the pipeline as pipe.start() without arguments (by default it should run on services queue as I understand right), then the clearm runs the pipeline, and in the terminal I see that it starts docker container… but nothing happens after I see: «you should upgrade pip»… so I abort this execution after an hour… may be you could advice a way to run services queue or some toy-example of code? I see an example in the GitHub, but it doesn’t help me, maybe I do something wrong… I have my own clearml server that I install ed with docker-compose 🤔

  
  
Posted 2 months ago

And if I will want to stop these queues, how should I do this ?

  
  
Posted 2 months ago

you will get PC:01 for a queue and PC:02 for other queue

  
  
Posted 2 months ago

Can we run a services queue for a pipeline and a default queue for tasks on the same machine?

  
  
Posted 2 months ago

open one more queue. one queue for pipeline another for actual task

  
  
Posted 2 months ago

run the same cmd twice

  
  
Posted 2 months ago