Host ai chat application with Axum: worker execution and job state

Summary

Deploy a ai chat application built with Axum as a public web service plus a dedicated worker process on Ample. ample plan --write discovers both services and their shared database and writes ample.toml; ample up reconciles the topology: a managed PostgreSQL 16 database, a public web microVM and a private worker microVM (kind worker, no public URL), both reading DATABASE_URL. Verified on Axum: a separately deployed worker service (kind worker, no public URL) sharing a declared managed PostgreSQL database with the web service, consuming a jobs table with row locks and marking jobs done once (worker=done attempts=1). Not separately tested: your job payloads, retry policy and scheduling; treat the ai chat application-specific behavior as your application code.

Infrastructure requirements

• Confirmed isolated x86_64 Firecracker microVMs for computing.
• Verified managed PostgreSQL 16 runs in its own microVM and is auto-provisioned when needed.

Prerequisites

• A repository with the web app and the worker as separate service directories.
• Both services read DATABASE_URL at runtime and share one jobs table.
• An Ample account token with servers:write and databases:write.

Exact tested configuration

• template: ubuntu-24.04
• runtime: rust
• size: s-1vcpu-2gb
• build: cargo build --release
• start: ./$(ls target/release/*.d | grep -v '/lib' | head -n 1 | sed 's/\.d$//')

Steps

1. Plan the topology. Run the planner once.

   ample plan --write .
   

2. Share one database. Keep a single [databases.main] with engine = "postgres".

   ample plan --offline .
   

3. Apply. Reconcile the manifest in dependency order.

   ample up .
   

4. Verify. Fetch the web service URL and its worker status route.

   ample logs <deployment_id> --kind runtime
   

Tested examples

• Axum web + worker fixture. A Axum web app and a worker process sharing one managed PostgreSQL database.

Success checks

• Web service responds on its public URL.
• Worker consumed the enqueued job.

Limitations

• Verified on the ubuntu-24.04 template at s-1vcpu-2gb for both services.
• The ai chat application’s behavior and specifics were not separately tested.
• The worker is a single long-running process with no horizontal scaling.

Cost estimate

Estimated 15.00 USD per month.

• web server x1 s-1vcpu-1gb: 5.00 USD
• worker server x1 s-1vcpu-1gb: 5.00 USD
• managed PostgreSQL database x1 s-1vcpu-1gb: 5.00 USD

Verification evidence

• canary_run on 2026-09-21T02:46:37Z: Axum web service plus a separate Axum worker process deployed on Ample.

Last verified: 2026-09-21T02:46:37Z

Execution binding

MCP tool ample_up, schema hash db7ca0facfe2cd28.