synchronous replication
Replication in which a write is only acknowledged after another copy has durably stored it.
In more detail
It protects committed writes against losing one machine, at the cost of extra latency and a new dependency: if the other copy is slow, every write is slow. Databases often let you choose it per transaction or per replica.
Examples
ship(primary, near) # synchronous: before we say "paid"Primary fails at 14:37: the synchronous replica still holds all 169 orders.Related terms
- asynchronous replication — Replication in which a write is acknowledged first and copied to other machines afterwards.
- round-trip time — The time it takes for a message to travel from one computer to another and for the reply to come back.
- RPO — Recovery point objective, the most data a business can afford to lose in an outage, measured as the time before the failure.
- geographic dispersion — Placing a primary site and its recovery site far enough apart that one local disaster cannot knock out both.
- replication — Keeping copies of data in more than one place, updated automatically, so one failure does not lose or block it.
- durability — The likelihood that stored data survives hardware failures without being lost, usually stated as a percentage per year.
Taught in
- Disaster Recovery Sites: Hot, Warm & Cold — Cybersecurity Fundamentals, GRC & Cryptography
- Backup Architectures: 3-2-1, Immutability & Air-Gaps — Cybersecurity Fundamentals, GRC & Cryptography
- Replication, Failover And Leader Election — Scalable System Design & Distributed Architecture