Read-your-writes consistency
A session guarantee: after a client writes, its own later reads see that write or a newer one, even if other clients still see old data. It matters when reads go to replicas that lag behind the primary; a common fix sends a user's reads to the primary for a short time after they write. Terry and others defined it in 1994.
Data
In Quorum#
A Read Replica lags behind the primary, so some reads come back stale, and a stale read fails an API read or a checkout that reads its own write. With Read-your-writes routing researched, turn the option on an App Server: checkout reads go to the primary and are never stale, and other stale reads fall by 80%. It also makes the Replication lag Incident much rarer (weight 0.3 instead of 2).
Read next#
Replication lag, Read replica, Eventual consistency, Synchronous replication