Skip to content

Running some services is too slow #1723

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

Closed
GitHK opened this issue Aug 20, 2020 · 1 comment · Fixed by #1734
Closed

Running some services is too slow #1723

GitHK opened this issue Aug 20, 2020 · 1 comment · Fixed by #1734
Assignees

Comments

@GitHK
Copy link
Contributor

GitHK commented Aug 20, 2020

Notes

This issue might be related with how we currently handle some things inside the services themselves. I will describe the scenario first.

Manifesting this issue is the sleeper service. All computational services should have the same issue as well. The dynamic services might inherit the same problem.

What actually happens

I was able to observe this happening only on master.

How to reproduce:

  • create a new study
  • add the service sleeper not the gpu or mpi version of it
  • press run button and look at the service logs

It will try to change permissions to around 5000 files in the input folder and 300 files in the output folder. These operations take a lot of time.

If you start a new terminal in the running service you can see there are a lot of directories inside the input folder.

Screenshot 2020-08-20 at 10 41 16

@GitHK
Copy link
Contributor Author

GitHK commented Aug 20, 2020

@pcrespov @sanderegg you might have more insight on why this is happening

@sanderegg sanderegg self-assigned this Aug 21, 2020
sanderegg added a commit that referenced this issue Aug 25, 2020
Sidecar mounts subfolder in computational service
* fix issue #1723 
* fix issue #1473
* diverse fixes and type annotations
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants