You saw in this chapter how to build slaves: the classical and GTID method. You now know how to optimize them, how to load balance against multiple read instances, and how to correct errors!

So much information in this chapter! It's strongly recommended to spend time to test every scenario, try to break replications, and so on, because in production it will not be so easy.

Replication is the first step of building a highly scalable system. That's why it's important to master it.

In the next chapter, we'll see how to implement those replications with more constraints, such as WAN architectures.