Multiple ChoiceIntermediate 45s
Read-your-writes consistency
To guarantee read-your-writes for a user after an update, which technique is commonly used with read replicas?
Options
- 1Force all reads to the leader for a short window (e.g. sticky routing or a version/LSN token)
- 2Add more read replicas so at least one replica has already applied the write by the time the user issues their read
- 3Switch replication from async to sync across all replicas forever, so no replica can ever lag behind the leader
- 4Increase the replica's connection pool so it serves reads faster than the write can propagate to it from the leader
Think you can answer under pressure?
Get matched and answer live questions in 3-minute 1v1 battles. Elo ranking, badge tiers and bragging rights on the line.