Skip to content

[Layer 2] Should the task DAG have depth/breadth limits to prevent resource exhaustion? #1

Description

@marc-poljak

Context. Layer 2 § Open Questions asks:
"What is the maximum DAG depth/breadth? Should the protocol define limits to prevent resource exhaustion?"

A task DAG that is arbitrarily deep or wide is a denial-of-service vector: a buyer could submit a graph that costs a provider large resources just to parse and schedule.

Your task. No implementation needed — this is an analysis. Consider:

  • Do comparable systems (e.g. Airflow, Argo Workflows, AWS Step Functions) impose depth/breadth limits? What values, and why?
  • Should ACMP mandate a limit, recommend a default, or leave it to each provider to advertise?
  • If advertised per-provider, which layer carries that field (L5 discovery? the L6 offer?)?

Done = a comment with your reasoning, or a PR editing the Open Questions section. A partial answer is a mergeable contribution — see CONTRIBUTING.

Metadata

Metadata

Assignees

No one assigned

    Labels

    good first issueGood for newcomersquestionFurther information is requestedwg-task-formatLayer 2 — Task Decomposition Format working group

    Type

    No type

    Fields

    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