Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Workflow invocations state for successful invocations and embedded #15428

Closed
itisAliRH opened this issue Jan 27, 2023 · 5 comments
Closed

Workflow invocations state for successful invocations and embedded #15428

itisAliRH opened this issue Jan 27, 2023 · 5 comments

Comments

@itisAliRH
Copy link
Member

Describe the bug

  1. All successful workflow invocations have scheduled state
    image
  2. Embedded workflow invocations state are unknown when parent workflow is scheduled, cancelled or failed.
    image
  3. Invocation state in the table shows the initial state when the list loaded not the last one
    image

Galaxy Version and/or server at which you observed the bug
usegalaxy.org (v23.0.rc1)

To Reproduce
Steps to reproduce the behavior:

  1. Go to Workflow Invocations from the User menu
  2. See successful/cancelled/failed invocations state

Expected behavior
Show successful for successful invocations and show the state of embedded workflow invocations based on their parent state and show the last state of each in the table

@mvdbeek
Copy link
Member

mvdbeek commented Jan 27, 2023

  1. All successful workflow invocations have scheduled state

I don't think that's a bug, they are scheduled. That has nothing to do with job states, so that is the only thing that really makes sense there. You can easily have failed jobs in a "successful" invocation. Are you maybe looking for an additional count of how many jobs are ok/paused/errored etc ?

2. Embedded workflow invocations state are unknown when parent workflow is scheduled, cancelled or failed.

Can you elaborate what you've done to reach that state ?

3. Invocation state in the table shows the initial state when the list loaded not the last one

Yes, there is no polling for updates. Possibly the reason for 2 ?

@itisAliRH
Copy link
Member Author

  1. All successful workflow invocations have scheduled state

I don't think that's a bug, they are scheduled. That has nothing to do with job states, so that is the only thing that really makes sense there. You can easily have failed jobs in a "successful" invocation. Are you maybe looking for an additional count of how many jobs are ok/paused/errored etc ?

I thought The "scheduled" state refers to an invocation that has not yet run and is waiting to run, and it would be appropriate to have another final state, such as "completed" or "finished," to indicate that the invocation has been completed regardless of the result.

  1. Embedded workflow invocations state are unknown when parent workflow is scheduled, cancelled or failed.

Can you elaborate what you've done to reach that state ?

One of the failed invocations failed for an unknown reason and the other one failed due to it having wrong dependencies (each image shows a workflow and the workflow embedded in it):
image
image
I have cancelled an invocation just after running it by clicking on the x button
image

  1. Invocation state in the table shows the initial state when the list loaded not the last one

Yes, there is no polling for updates. Possibly the reason for 2 ?

I think yes

@mvdbeek
Copy link
Member

mvdbeek commented Jan 27, 2023

Scheduled means the invocation has been scheduled successfully, that means all steps have been created in the database. If jobs for all steps can be scheduled ahead of time it's often fairly quick / immediate, but it can happen that we can't create jobs yet because no reference to intermediate datasets exist, in that case the scheduler is waiting until that is the case.

You've seen the other cases, an invocation can be cancelled (on user request or be deleting the history) and it can fail (for a variety of reasons, like output datasets that are never going to be produced, or if jobs fail that are required in order to proceed scheduling more jobs).

@mvdbeek
Copy link
Member

mvdbeek commented Jan 27, 2023

We should probably explain ^^ in a glossary or tooltip or something like that

@dannon dannon added this to the 23.1 milestone Jan 31, 2023
@mvdbeek mvdbeek modified the milestones: 23.1, 23.2 Jul 21, 2023
@mvdbeek mvdbeek removed this from the 23.2 milestone Jan 11, 2024
@mvdbeek
Copy link
Member

mvdbeek commented Jul 20, 2024

That actually happened
Screenshot 2024-07-20 at 09 06 42 and there's #17708 to make the two states easier to see.

@mvdbeek mvdbeek closed this as completed Jul 20, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

3 participants