Is there a downside to running a worker to dispatch to a kubernetes work pool in the same pod as the prefect server? I'm running on a GKE autopilot cluster that requires significant mimimum resources that should support both processes.

hi @Sean Davis - can you expand on this a bit? > running a worker to dispatch to a kubernetes work pool typically a worker runs in its own pod and picks up work from / gets a job infra template from a work pool

If I were on a non-autopilot cluster, I would definitely run the worker in its own pod. In the context of an autopilot cluster, I'm trying to avoid running two pods since the pod minimum requirements are significant for the workload (a few flows/day for the worker, 1 user for the server). This is purely about utilizing resources efficiently.

gotcha! > simple system since scaling won't be needed on either the server or the worker for this context this seems reasonable to me 👍 but im not a GKE autopilot expert 🙂