Dear Developers,
Is the Repeatable-Read isolation level based on the so-called 'Snapshot Isolation' by Hal Berenson's paper 'A Critique of ANSI SQL Isolation Levels'? If the answer is yes, is there a plan to implement the so-called 'Serializable Snapshot Isolation' by Michael Cahill's paper 'Serializable Isolation for Snapshot Databases' to replace it or the lock-based implementation of 'Serializable' isolation level. PostgreSQL has done that since release version 9.1 and I am wondering if innodb is going to do that too.
Cheers,
David
Is the Repeatable-Read isolation level based on the so-called 'Snapshot Isolation' by Hal Berenson's paper 'A Critique of ANSI SQL Isolation Levels'? If the answer is yes, is there a plan to implement the so-called 'Serializable Snapshot Isolation' by Michael Cahill's paper 'Serializable Isolation for Snapshot Databases' to replace it or the lock-based implementation of 'Serializable' isolation level. PostgreSQL has done that since release version 9.1 and I am wondering if innodb is going to do that too.
Cheers,
David