Rate-limit your app.
One bad actor — a brute-force script, a runaway client, a scraper — shouldn't be able to degrade your service for everyone. Set a limit and get an nginx config with burst handling, a 429 response, and an optional strict limiter for login routes.
One client shouldn't sink the ship
Most outages that look like attacks aren't sophisticated — they're one client in a retry loop, one scraper with no backoff, one brute-force script hammering your login. Without a limit, a single source can saturate your app and database and take the whole service down for everyone else. A few lines of rate-limiting cap each client to something reasonable and turn that from an outage into a handful of 429s.
Rate limiting at the proxy handles a configured origin-side case. A control plane can report selected abuse evidence and adjust enabled policy for managed routes, but it does not provide continuous awareness across every host or upstream network path.
Cap the abuse. See the rest.
For configured managed routes, Infraveil can report selected request and abuse signals and bounded origin-policy evidence. Coverage and retention depend on configuration; upstream volumetric attacks, saturated links, unmanaged hosts, and bypass listeners remain outside the managed gateway boundary.
See how it works