Crate orbit_runtime 

The ORBIT runtime: a durable, single-node workflow execution service.

• model — history events and the state folded from them
• decide — the pure, deterministic workflow decider
• store — SQLite persistence (event history is authoritative)
• engine — dispatch, worker registry, reconciliation and recovery
• server — HTTP API and the HTTP/JSON worker protocol binding
• cluster — active/standby operation of several servers on PostgreSQL
• shared — several active servers sharing the work on PostgreSQL
• testing — in-process simulation for orbit test and orbit simulate
• ui — the built-in web UI served at /ui/
• replay — re-run a recorded history through the current decider
• schedule — cron and interval schedules that start executions

Re-exports

pub use engine::Engine;

pub use engine::EngineConfig;

pub use engine::EngineError;

pub use store::Store;

Modules

cluster

Multi-server operation: one active leader, any number of standbys, all on the same PostgreSQL database (orbit server --cluster).

decide

The deterministic workflow decider.

engine

The execution engine: owns persistence, in-memory projections of active executions, the worker registry and task dispatch.

grpc

gRPC binding of the worker protocol (orbit.worker.v1.WorkerService).

http_adapter

Built-in orbit.http.request task: lets a workflow call an HTTP endpoint without writing a worker. Disabled unless the server is started with --enable-http-adapter, because it lets workflow authors make outbound requests from the runtime host.

identity

Per-worker identities.

metrics

Prometheus-style counters derived from committed events.

model

Execution history events and the state derived from them.

replay

Deterministic replay: re-run a recorded history through the current decider and check that every decision the runtime recorded is reproduced.

schedule

Schedules: workflows started by a cron expression or a fixed interval.

server

HTTP API (RFC-0001 §11.2) and the HTTP/JSON binding of the worker protocol (§8). See protocol/README.md for the wire format.

shared

Shared mode: several active servers on one PostgreSQL database (orbit server --cluster --cluster-mode shared). Every node serves every route; executions, worker tasks and worker sessions are spread over them.

store

Persistence.

telemetry

OpenTelemetry tracing: OTLP/HTTP (JSON encoding) export of execution traces, accepted by the OpenTelemetry Collector, Jaeger, Grafana Tempo, Honeycomb and other OTLP backends.

testing

In-process workflow simulation, used by orbit test and orbit simulate.

ui

The built-in operations UI (/ui/).