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

Inconsistency in accelerator information between session launcher and resource monitor #3188

Open
ironAiken2 opened this issue Dec 2, 2024 · 0 comments

Comments

@ironAiken2
Copy link

What Operating System(s) are you seeing this problem on?

Linux (x86-64)

Backend.AI version

24.09.2rc1

Describe the bug

Issue Summary

Inconsistency in accelerator information between session launcher and resource monitor can cause user confusion.

Description

The session launcher uses the /func/config/resource-slots/details?sgroup=multinode API, which only shows detailed information for the supported resources in the group. As a result, if an accelerator is not supported, it is not included in the response.

On the other hand, the resource monitor uses the Check-preset API, which includes accelerators in the response even if they are not supported by the resource group, as long as they have been reported to the manager at least once. In such cases, the total value is set to 0 to indicate that the accelerator is not supported.

This inconsistency can lead to user confusion.

Additional Information

To Reproduce

You can reproduce it in the Multinode resource group in dogbowl.

Expected Behavior

No response

Anything else?

No response

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

No branches or pull requests

1 participant