Cascading failure.
also: cascading failures · cascade failure · cascade failures · death spiral · thread pool exhaustionOne slow or failing service exhausts its callers' threads or connections, so failure spreads up the call chain.
the two-minute refresher
- C slows down; B's threads all wait on C; B stops accepting A's calls; A backs up too.
- Slow is worse than down: a dead service fails fast, a slow one holds resources.
- Defences: tight timeouts, circuit breakers, bulkheads and load shedding.
- Retries without backoff make it worse.