Agent queues vs graph runtimes
How to choose between a durable queue, a workflow graph, and a hybrid runner for agent work.
Queues and graphs solve different reliability problems. The queue gives durability and backpressure; the graph gives explicit control flow and state transitions.
AstraRunner treats this as an execution-platform concern: the run needs durable state, traceable decisions, cost visibility, and enough structure for teams to review what happened after the agent finishes.
Patterns that work
- Start with a queue when each task can complete independently.
- Move to a graph when the task branches on validation, approval, or tool output.
- Use a hybrid when queued work needs graph-shaped substeps.
- Keep graph nodes small enough to retry safely.
Implementation checklist
- Define the unit of retry.
- Store graph state outside the worker process.
- Track queue age and dead-letter counts.
- Expose a human view for stuck nodes.
Watch outs
- Graph diagrams can hide expensive fanout.
- Queues without idempotency duplicate side effects.
- In-memory graph state breaks on deploys and worker crashes.
Related reading
How this fits the Astra stack
AstraRunner owns the operational path for agent work: orchestration, scheduling, approval gates, traces, costs, and role-agent handoffs. AstraMemory owns durable context and retrieval. AstraGenie turns those capabilities into product workflows for teams that want automation without rebuilding the platform layer.