What safety or application-level bugs could arise if developers assume Snapshot Isolation but Amazon RDS for PostgreSQL is actually providing only Parallel Snapshot Isolation, especially in multi-AZ configurations using the read replica endpoint?
And the mere fact that it's hard to reason about these things means that it's hard to avoid problems. Hence, the easiest solution is likely "it may be possible to recover Snapshot Isolation by only using the writer endpoint", for anything where write is anyhow conditional on a read.
Although I'm surprised the "only using the writer endpoint" method wasn't tested, especially in availability loss situations.
- User1 comments
- User2 comments
- User1 checks (in a separate tx) that there's only 1 comment, so User1 gets the badge
- User2 checks the same (in a separate tx) and also sees only 1 comment (his), and also receives the badge.
With Snapshot isolation this isn't possible. At least one of the checks made in a separate tx would see 2 comments.
The original article on the Parallel Snapshot is a good read: https://scispace.com/pdf/transactional-storage-for-geo-repli...