Ployz: cluster orchestration built on explicit primitives
Ployz turns cluster operations into explicit, atomic commands. Instead of writing manifests and waiting for reconcilers, you run ployzctl deploy, ployzctl branch, or ployzctl migrate — each command completes or fails cleanly, every time. Ployz targets clusters in the 1–200 node range: small enough to reason about end-to-end, large enough to matter.
Go from a fresh machine to a running cluster in under five minutes.
Install ployzctl and the daemon on Linux or macOS.
Understand what makes Ployz different from Kubernetes and other orchestrators.
Every ployzctl command, flag, and option, documented with examples.
What Ployz does
Section titled “What Ployz does”Ployz gives you a small set of strong operations — primitives — that compose into any workflow you need. There are no controllers, no autoscalers, and no background reconcilers silently rewriting cluster state.
Add and remove machines. Drain workloads before removal.
Deploy workloads from a manifest or inline flags. Preview before applying.
Fork an environment atomically for a PR, then promote it to production.
Move a workload and its persistent state to a different machine.
How it works
Section titled “How it works”Install ployz
Section titled “Install ployz”Run the one-line installer to get ployzctl and the ployzd daemon on your machine.
curl -fsSL https://ployz.sh | bash -s -- installCreate a cluster network
Section titled “Create a cluster network”Initialize a WireGuard mesh network that your machines will join.
ployzctl mesh init my-clusterAdd machines
Section titled “Add machines”Provision additional machines into your cluster with a single command.
ployzctl machine add --network my-cluster user@192.168.1.10Deploy your first workload
Section titled “Deploy your first workload”Deploy a container image to your cluster.
ployzctl deploy -f deploy.toml