Supporting Services
The Service Map catalogues the trading pipeline itself (market sim, OMS, EMS, algos, journal, FIX, and the analytics services). This section covers the supporting infrastructure that surrounds it: ingress, load testing, and operational utilities. Metrics, logs and traces have moved to their own Observability page.
Traefik Ingress proxy that fronts every backend service.
Synthetic probe External 60s liveness check from the edge server, with a CI mirror after every merge.
veta-auto-pull Systemd timer that polls origin/main every 5 min and deploys when a new commit lands. The server's continuous deployment.
veta-tunnel Secure tunnel that exposes the server publicly. The server dials out, so no inbound NAT is required.
veta-host-prune Weekly Docker prune timer. Prevents image churn from filling the server disk.
k6 load testing On-demand synthetic load harness for single runs against a local stack.
Continuous load generator Always-on trade flow on the production server: soak and matrix runners with OAuth refresh.
Disk monitor Read-only host-disk gauge exposed at /metrics for Prometheus scraping; healthcheck-decoupled so a full disk surfaces as an alert, not an unhealthy container.
db-migrate One-shot Postgres migration runner.
Redpanda console Optional Kafka UI for debugging the message bus.
Where to run what
Section titled “Where to run what”| Environment | Trading stack | LGTM stack | k6 |
| --- | --- | --- | --- |
| local (devcontainer) | yes | optional | optional |
| production (https://veta.mnetcs.com) | yes | yes | manual |
LGTM and k6 are optional locally. They add real overhead and the platform runs fine without them. Bring them up when investigating performance, debugging event flow, or generating data for the performance docs.