Skip to content

Client to a cluster with more than 5 workers appears as though it only has 5 #9065

@zmbc

Description

@zmbc

Describe the issue: Clients only report information about the cluster's first five workers without a disclaimer.

Minimal Complete Verifiable Example:

from distributed import LocalCluster, Client

if __name__ == '__main__':
    cluster = LocalCluster(n_workers=6, memory_limit="500MiB")
    c = Client(cluster)
    print(c)

Output is <Client: 'tcp://127.0.0.1:46429' processes=5 threads=5, memory=2.44 GiB> when it should be processes=6 threads=6, memory=3 GiB

Anything else we need to know?: This change introduced this issue. cc @fjetter

Environment:

  • Dask version: 2025.4.1
  • Python version: 3.11.9
  • Operating System: Linux
  • Install method (conda, pip, source): pip

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething is broken

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions