fix: correct PCSG topology constraint handling for scaled PodGangs#357
Merged
unmarshall merged 9 commits intoJan 20, 2026
Merged
Conversation
Signed-off-by: Ron Kahn <rkahn@nvidia.com>
Add nil check before creating PCSG topology constraint groups in base PodGang building. This prevents creating empty topology constraints when PCSG has no TopologyConstraint specified, ensuring scaled PodGangs properly fall back to PCS-level constraints. Signed-off-by: Ron Kahn <rkahn@nvidia.com>
… pcs Signed-off-by: Ron Kahn <rkahn@nvidia.com>
Signed-off-by: Ron Kahn <rkahn@nvidia.com>
Signed-off-by: Ron Kahn <rkahn@nvidia.com>
…_test.go Signed-off-by: Ron Kahn <rkahn@nvidia.com>
…t.go Signed-off-by: Ron Kahn <rkahn@nvidia.com>
Signed-off-by: Ron Kahn <rkahn@nvidia.com>
shayasoolin
approved these changes
Jan 20, 2026
unmarshall
approved these changes
Jan 20, 2026
danbar2
pushed a commit
to danbar2/grove
that referenced
this pull request
Jan 21, 2026
…i-dynamo#357) * Revert "fix: add PCSG topology constraints to scaled PodGangs (ai-dynamo#347)" This reverts commit a08e5af. Signed-off-by: Ron Kahn <rkahn@nvidia.com> * fix: update topology constraint handling for scaled PodGangs Signed-off-by: Ron Kahn <rkahn@nvidia.com> * fix: only create PCSG topology constraints when defined Add nil check before creating PCSG topology constraint groups in base PodGang building. This prevents creating empty topology constraints when PCSG has no TopologyConstraint specified, ensuring scaled PodGangs properly fall back to PCS-level constraints. Signed-off-by: Ron Kahn <rkahn@nvidia.com> * test: add case for PCS with nil PCSG topology constraints fallback to pcs Signed-off-by: Ron Kahn <rkahn@nvidia.com> * test: clean up formatting in syncflow_test.go Signed-off-by: Ron Kahn <rkahn@nvidia.com> * test: clean up formatting in syncflow_test.go Signed-off-by: Ron Kahn <rkahn@nvidia.com> * test: remove redundant comments regarding pcsgConstraints in syncflow_test.go Signed-off-by: Ron Kahn <rkahn@nvidia.com> * test: improve formatting and remove redundant comment in syncflow_test.go Signed-off-by: Ron Kahn <rkahn@nvidia.com> * test: clean up formatting in syncflow_test.go Signed-off-by: Ron Kahn <rkahn@nvidia.com> --------- Signed-off-by: Ron Kahn <rkahn@nvidia.com>
Ronkahn21
added a commit
to Ronkahn21/grove
that referenced
this pull request
Jan 21, 2026
- Rename test functions to simplified format (TAS1-TAS7) - Remove scenario labels from test comments - Replace hardcoded strings with setup package constants - Update log messages to use new test names - Fix PR ai-dynamo#357 semantic change: PCSG inherits PCS constraint when nil - Remove duplicate topology functions from k8s_clusters.go Signed-off-by: Ron Kahn <rkahn@nvidia.com>
Ronkahn21
added a commit
to Ronkahn21/grove
that referenced
this pull request
Jan 21, 2026
When PCSG has nil TopologyConstraint, PCSG parent groups are not created in KAI PodGroup. PCLQ children are placed directly under PCS-level constraint instead. - Test_TAS3: Changed from 5 to 3 SubGroups (removed PCSG parents) - Test_TAS7: Changed from 4 to 2 SubGroups (removed PCSG parents) - Updated PCLQ children to have Parent: nil instead of PCSG parents Signed-off-by: Ron Kahn <rkahn@nvidia.com>
Ronkahn21
added a commit
to Ronkahn21/grove
that referenced
this pull request
Jan 24, 2026
- Rename test functions to simplified format (TAS1-TAS7) - Remove scenario labels from test comments - Replace hardcoded strings with setup package constants - Update log messages to use new test names - Fix PR ai-dynamo#357 semantic change: PCSG inherits PCS constraint when nil - Remove duplicate topology functions from k8s_clusters.go Signed-off-by: Ron Kahn <rkahn@nvidia.com>
Ronkahn21
added a commit
to Ronkahn21/grove
that referenced
this pull request
Jan 24, 2026
When PCSG has nil TopologyConstraint, PCSG parent groups are not created in KAI PodGroup. PCLQ children are placed directly under PCS-level constraint instead. - Test_TAS3: Changed from 5 to 3 SubGroups (removed PCSG parents) - Test_TAS7: Changed from 4 to 2 SubGroups (removed PCSG parents) - Updated PCLQ children to have Parent: nil instead of PCSG parents Signed-off-by: Ron Kahn <rkahn@nvidia.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What type of PR is this?
/kind bug
What this PR does / why we need it:
Fixes incorrect PCSG topology constraint handling for scaled PodGangs:
Which issue(s) this PR fixes:
Fixes #339
Special notes for your reviewer:
Does this PR introduce a API change?
Additional documentation e.g., enhancement proposals, usage docs, etc.: