Skip to content

K8s readiness / liveness probes for workers #141

@thockin

Description

@thockin

Do we need readiness and liveness probes for workers?

It seems important to have readiness of the worker itself, but if we forward all traffic into the actor, does that mean all actors must implement our readiness protocol? Or does it mean we have to use exec or something (not network)?

How about liveness? It seems like we should be able to detect a locked-up agent -- does that mean all actors must implement our liveness protocol?

Metadata

Metadata

Assignees

No one assigned

    Labels

    area/apiUser-facing API changesarea/nodeprio/P1Important but not critical

    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