Architecture · 6 nodes
Distributed basics
What the network does to assumptions you did not know you had made, until it breaks them. Retries, ordering, and the honest version of exactly-once.
Is this path right for me?
Nothing is tracked and nothing is locked — the map exists so you can see where an article sits.
Suggested start Written Not written yet Optional or subtopic
-
What the network can do to you
- Partitions
- Timeout is not failure
-
Retries and idempotency
See why retries hit hardest exactly when a system is already failing, why spacing them out is not enough on its own, and where the multiplication comes from.
- Backoff and jitter
- Dedupe windows
-
Ordering
- Per-key order
- Causality
-
Exactly-once, honestly
- Effectively-once
- The outbox pattern
-
Consensus, from a distance
- Quorums
- Leases
-
Observability when it breaks
See what one header has to carry for a request to stay one story across a dozen services, and why the identifiers are standardised but the meaning is not.
- Trace context
- Correlation ids