I assume we use 'row-based replication'
Jeff_Green::: eileen: something occured to me over the weekend: triggers exist only on the master right?
eileen:: Jeff_Green: hmm - maybe they get replicated. That would be bad
10:30 eileen: i was thinking about what happens if the master explodes
10:30 Jeff_Green: wrt triggers, yeah I think that's the interaction that causes all of our SUPER problems, cos triggers and functions on the replica would not be deterministic
10:30 we need to be sure we have a cutover protocol that allows us to promote a slave without losing triggers
10:31 ooh good catch
10:31 hmm - as long as the triggers don't fire during replication
10:31 it woke me up actually
10:31 ha, a demonic vision
10:31 just about to fall asleep panic on Saturday night
10:32 yeah we can't really test that on staging!
10:32 i might be able to figure out a way to test it
10:32 am just reading this https://dev.mysql.com/doc/refman/5.7/en/replication-features-triggers.html