What is the best way to "mirror" MySQL data on two servers?

Question

linux

Hello!
I have a very strange question =)
I decided to organize MySQL replication according to the Master - Slave scheme , for "mirroring" data on two servers.
There are " production " and " reserved " sites in different DCs.
In case of " production " crash, we go to work (manually) on the " reserved " site.
For implementation, I chose replication and in case of failure on " production " employees continue to work with the same data on the " reserved " site.
I was advised to do this: if we switched to work on "", then it works with this database until the end of the working day. At the end of the day, we manually synchronize with " production " and start replication. And accordingly, we can work with " production " again . by default, but as soon as it has fallen, we turn on " reserved ". The essence of the question is: how best to implement a stub for " production ", if it "came to life" after a fall, for example, after an hour and in order to avoid different data on MySQL servers, disable access to it at the Nginx level, eg ?


" and be activated manually. Then the next day, for example, the stub will be turned on again.
In general, can you advise in which direction it is better and easier to dig, because you want to achieve uninterrupted operation of only MySQL and the relevance of the data on two servers. And in view of this, I am trying dance with different tambourines and crutches =)
OS: Ubuntu 14.04
MySQL: 5.5.40
Database size: about 80MB

Answer 1

www.slideshare.net/tsarevoleg/ss-40969331

Answer 2

https://ru.wikipedia.org/wiki/%D0%A2%D0%B5%D0%BE%D...