OTP Supervision Trees
← Back to Erlang-Elixir Concurrency
A hierarchical structure where supervisor processes monitor worker processes. When a worker crashes, its supervisor restarts it according to a configured strategy (one-for-one, one-for-all, rest-for-one). This “let it crash” philosophy provides fault tolerance through isolation and automatic recovery.