Docs
Infrastructure requirements

Infrastructure requirements

Plan capacity and dependencies before running AITracer in production.

What you need depends on trace volume, retention, concurrency, and which product features you use. This page summarizes requirements and sizing considerations—not a prescribed network or server layout.


What you run

AITracer ships as a web application with a backing PostgreSQL database. In production you typically also place HTTPS (TLS) in front of the app and manage secrets outside the repo (environment variables or a secret manager).

Optional capabilities may increase needs:

  • Long retention or large exports — extra database storage or separate object storage you control
  • High ingest rates — more CPU and memory for the app tier; scale horizontally behind your load balancer as you would any stateless web service

Compute

Plan for:

  • the application runtime (container or VM) sized for your peak API and dashboard traffic
  • headroom for background work the product performs during ingest and governance

Use autoscaling if your traffic is spiky; there is nothing AITracer-specific beyond ordinary web-app scaling practices.


Storage

Growth is driven mainly by:

  • how long you keep traces and related metadata
  • prompt/response payload sizes you choose to retain
  • audit and verification artifacts you store

PostgreSQL is the primary datastore. If you archive or export to cold storage, size that according to your compliance and retention policy.


Networking

From the outside:

  • HTTPS to the application
  • DNS pointing at your entry point

From your apps to AITracer:

  • reachability to your trace ingestion and API base URL
  • firewall rules as required by your security team

Private networking, VPNs, and regional routing are your operational choices—not part of this product’s public contract.


Observability

Run the app with your standard logging, metrics, and alerting so you can see errors, latency, and saturation. AITracer does not mandate a particular vendor.


Capacity checklist

Use this when planning:

  • expected requests per day and peak concurrent users
  • retention duration and whether payloads are stored in full or summarized
  • compliance storage and export obligations

For configuration keys and auth, see Environment configuration.