Throughout my career I have faced the challenge of issues with MySQL’s built-in replication. The biggest problem has always been in regards to the consistency of a replica with its master.
I know that network issues, MySQL bugs, crashes, and even MySQL features can cause inconsistencies in data. In fact, my previous employer had a huge problem with this. But I was had a hell of a time trying to figure out why.
I had stopped trying to figure out why a while back and started to use Percona’s scripts to detect and resolve. But after reading the most recent edition of “High Performance MySQL” my desire to know more about this phenomenon has increased.
The book talks about “data drift”, but doesn’t go into a lot of detail about it. Does anyone know of any whitepaper or blogs that go in to more depth about this issue?