kubeflow/trainer

The helpers RuntimeInfo helpers don't handle correctly numNodes

Open

#3,042 opened on Dec 17, 2025

View on GitHub
 (10 comments) (0 reactions) (2 assignees)Go (964 forks)github user discovery
area/controllergood first issuehelp wantedkind/bug

Repository metrics

Stars
 (2,112 stars)
PR merge metrics
 (PR metrics pending)

Description

What happened?

We are using the Kubeflow Trainer helpers in Kueue to construct the PodSets.

However, when numNodes >0 we need to post-process the results to get the proper numbers of Pods.

See the bugfix: https://github.com/kubernetes-sigs/kueue/pull/8135/changes#diff-eb30173c6ca84a779cd005d1c147a311b18a077bd35205dc64942fd824ffb3e4R181-R187

Here is the follow up issue in Kueue: https://github.com/kubernetes-sigs/kueue/issues/8287

What did you expect to happen?

I would expect the outcome of the helpers not to require any custom post-processing. Other projects may also hit this issue.

Environment

Kueue project

Impacted by this bug?

Give it a 👍 We prioritize the issues with most 👍

Contributor guide

Research direction
Investigate the helper functions in the RuntimeInfo area to understand how numNodes is processed and ensure that the number of Pods is correctly calculated without requiring post processing.
Tech stack
go
Domain
backend
Issue type
Bug
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
2
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
1-3 hours
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Clear
Prerequisites
Go
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
60