How We Told a Stranger's Node Where Its Cache Should Be

We connected to a remote BEAM node we don't own. No access to its source code. No instrumentation planted beforehand. No agents installed, no probes compiled into the target. Pure black-box runtime observation over Distributed Erlang. From that, we produced a concrete architectural recommendation: your schema registry is hitting PostgreSQL 354 times per observation window for metadata that almost never changes — that should be ETS, not a database table . This article walks through how we got there — 7 observation sessions against a live Elixir application, progressing from coarse process-level metrics to function-level tracing that revealed the internal structure of code we'd never read. The Target: Nexus Nexus is an Elixir API gateway that manages dynamic PostgreSQL connections, runtime schema metadata, and table operations. It exposes a REST API on port 4040 and handles CRUD, aggregations, search, CSV export, batch operations, and multi-database routing — all backed by Ecto dynamic r