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
I Dont Exactly Know How To Ask For Help On This... Nor Have A Reproducible Minimal Example... I Downgraded Back To 1.15.1 From 1.16.2 And Have The Same Issue There. I Have A Pipeline That'S Repeatedly Failing To Complete. It Correctly Marks Things As Cach

i dont exactly know how to ask for help on this... nor have a reproducible minimal example...
I downgraded back to 1.15.1 from 1.16.2 and have the same issue there.
I have a pipeline that's repeatedly failing to complete. it correctly marks things as cached, and then just doesnt execute the last step. The task stays "Running" forever, but disappears - the worker just has a process that dies. CPU/RAM aren't running out or anything like that at all. Prior runs of this pipeline worked just fine (less cached then too).

Has anyone seen this kind of "disappearing / zombie task" state before? It's very perplexing to me.

  
  
Posted 3 months ago
Votes Newest

Answers 43


its odd... I really dont see tasks except the controller one dying

  
  
Posted 3 months ago

trying to run the experiment that kept failing right now, watching logs (they go by fast)... will try to spot anything anamolous

  
  
Posted 3 months ago

the workers connect to the clearml server via ssh-tunnels, so they all talk to "localhost" despite being deployed in different places. each task creates artifacts and metrics that are used downstream

  
  
Posted 3 months ago

N/A (still shows as running despite Abort being sent)

  
  
Posted 3 months ago

yeah locally it did run. I then ran another via UI spawned from the successful one, it showed cached steps and then refused to run the bottom one, disappearing again. No status message, no status reason. (not running... actually dead)
image

  
  
Posted 3 months ago

do you have any STATUS REASON under the INFO section of the controller task?

  
  
Posted 3 months ago

ah. a clue! it came right below that but i guess out of order...
that id is the pipeline that failed
image

  
  
Posted 3 months ago

Hi @<1689446563463565312:profile|SmallTurkey79> ! I will take a look at this and try to replicate the issue. In the meantime, I suggest you look into other dependencies you are using. Maybe some dependency got upgraded and the upgrade now triggers this behaviour in clearml.

  
  
Posted 3 months ago

would it be on the pipeline task itself then, since that's what's disappearing?
I will do some experiment comparisons and see if there are package diffs. thanks for the tip.

  
  
Posted 3 months ago

clearml_agent v1.8.1
  
  
Posted 3 months ago

yeah, it just shows what I see in the Console, but then immediately goes back to polling for more work (so... instead of running backtest, it exits, no completion message)

  
  
Posted 3 months ago

clearml-server-1.15.1, clearml-1.16.2
  
  
Posted 3 months ago

I have tried other queues, they're all running the same container.
so far the only thing reliable is pipe.start_locally()

  
  
Posted 3 months ago

odd bc I thought I was controlling this... maybe I'm wrong and the env is mis-set.
image
image

  
  
Posted 3 months ago

did you take a look at my connect.sh script? I dont think it's a problem since only the controller task is the problem.

Is there some sort of culling procedure that kills tasks by any chance? the lack of logs makes me think it's something like that.

I can also try different agent versions.

  
  
Posted 3 months ago

default queue is served with (containerized + custom entrypoint) venv workers (agent services just wasn't working great for me, gave up)

  
  
Posted 3 months ago

hoping this really is a 1.16.2 issue. fingers crossed. at this point more pipes are failing than not.

  
  
Posted 3 months ago

yeah this problem seems to happen on 1.15.1 and 1.16.2 as well, prior runs were on the same version even. It just feels like it happens absolutely randomly (but often).
just happened again to me.

The pipeline is constructed from tasks, it basically does map/reduce. prepare data -> model training + evaluation -> backtesting performance summary.

It figures out how wide to go by parsing the date range supplied as input parameter. Been running stuff like this for months but only recently did things just start... vanishing like this.

Would appreciate any help. Really need this to be more robust to make the case for company-wide adoption.
image

  
  
Posted 3 months ago

it happens consistently with this one task that really should be all cache.
I disabled cache in the final step and it seems to run now.

  
  
Posted 3 months ago

it's pretty reliably happening but the logs are just not informative. just stops midway
image

  
  
Posted 3 months ago

do you have the agent logs that is supposed to run your pipeline? Maybe there is a clue there. I would also suggest to try enqueuing the pipeline to some other queue, maybe even run the agent on your on machine if you do not already and see what happens

  
  
Posted 3 months ago

thank you

  
  
Posted 3 months ago

Hi @<1689446563463565312:profile|SmallTurkey79> !
Prior runs of this pipeline worked just fine What SDK version were you using for the prior runs? Does this still happen if you revert to that version?
Can you provide a script that imitates what you are doing?
In the pipeline you are running, are you creating new tasks/pipelines/datasets?

  
  
Posted 3 months ago

damn, it just happened again... "queued" steps in the viz are actually complete. the pipeline task disappeared again without completion, logs mid-stream.

  
  
Posted 3 months ago

when i run the pipe locally, im using the same connect.sh script as the workers are in order to poll the apiserver via the ssh tunnel.

  
  
Posted 3 months ago

I will ask internally about this

  
  
Posted 3 months ago

but maybe here's a clue. after hanging like that for a while... it seems like the agent restarts (the container it runs in does not)
image

  
  
Posted 3 months ago

would it be on the pipeline task itself then, since that's what's disappearing? that likely the case

  
  
Posted 3 months ago

image

  
  
Posted 3 months ago

I really can't provide a script that matches exactly (though I do plan to publish something like this soon enough), but here's one that's quite close / similar in style:
None where I tried function-steps out instead, but it's a similar architecture for the pipeline (the point of the example was to show how to do a dynamic pipeline)

  
  
Posted 3 months ago
10K Views
43 Answers
3 months ago
3 months ago
Tags