SCHED-1648: Better support for Block topology - #2567
Merged
Merged
Conversation
dstaroff
marked this pull request as ready for review
June 2, 2026 09:29
asteny
reviewed
Jun 2, 2026
asteny
reviewed
Jun 2, 2026
asteny
reviewed
Jun 2, 2026
asteny
reviewed
Jun 2, 2026
ali-sattari
reviewed
Jun 2, 2026
ali-sattari
approved these changes
Jun 2, 2026
asteny
approved these changes
Jun 2, 2026
ali-sattari
pushed a commit
that referenced
this pull request
Jun 17, 2026
SCHED-1648: Better support for Block topology
andriishestakov
pushed a commit
to andriishestakov/soperator
that referenced
this pull request
Jul 8, 2026
SCHED-1648: Better support for Block topology
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Problem
Existing
worker-initinit container dynamically registers worker topology as if the Slurm cluster is always hasTopologyPlugin=topology/tree- it generates worker topology as this:default:<tier0>:<tier1>:<tier2>:<worker>. This doesn't work for block topology, as it expects workers to be part of only one block.Solution
Provide separate dynamic topology generation based on the topology plugin:
topology/tree- nothing changestopology/block- generatedefault:<tier0>:<worker>Additional changes
Support for pattern names in
topology.confWe used to generate a comma-separated list of nodes, which may hit the CM size limit on really big clusters. Now we use patterns which are length-efficient and easy-readable. Sorted order and possible gaps are managed automatically.
Support for
topologysection of SlurmCluster CR to be set by HelmIt was impossible to set e.g.
blockSizevia Helm.Release Notes
Feature:
TopologyPlugin=topology/blockis supported for dynamic worker topology setting.Feature:
topology.confuses pattern names for nodes instead of comma-separated list.Feature:
topologysection is possible to change viaslurm-clusterHelm chart.