Reliability

What it is

Reliability means preserving correct outcomes despite expected failures.

When to use it

Design retries, timeouts, duplicates, and recovery deliberately.

Syntax

bounded timeout + retry-safe operation + durable state + reconciliation

Example

Provider timeout after sending payment:
1. Mark outcome unknown/pending.
2. Query provider by stable reference.
3. Do not blindly create another payment.
4. Reconcile later if still unknown.

Common mistakes

A timeout is not proof of failure; indefinite retries can amplify outages.

Related topics