Can Distributed Databases Achieve Transactional Consistency on Async Standbys? Yes, They Can.

A distributed database like YugabyteDB achieves transactional consistency on an async standby by reading at a consistent timestamp equal to the minimum caught up time across all tablets in the database, ensuring atomicity and global ordering of transactions. This is achieved through cross-cluster replication, which replicates data from the active cluster to the standby with some lag, but ensures that reads occur at a safe time below which no new records can come in, preventing transactional violations like atomicity and global ordering issues. The tradeoff for achieving this consistency is staleness of reads and potential data loss on failover corresponding to the lag between active and standby.