Skip to content

Schedulers and workers do not know their cluster provenance #5031

@jacobtomlinson

Description

@jacobtomlinson

When creating clusters via Cluster objects the provenance of that creation is not communicated downstream to each component.

If I create a cluster with

from dask.distributed import LocalCluster
cluster = LocalCluster()

And then inspect the cluster.scheduler object or create an RPC directly with the scheduler and have a poke around there is no reference to the cluster object that created it, the cluster's name or the fact that it was created by LocalCluster as opposed to KubeCluster.

In the longer-term goal to make cluster objects reconstructible (via tools like dask-ctl) it would be really useful if some information like the cluster ID and class is passed on to the scheduler.

This could also be helpful as part of #5012 as the scheduler could broadcast this information using mDNS which would aid discovery.

This could also tie in with #4607 where this information is placed into that store.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions