If a key is getting updated twice for a master slave
-The data is always sent to the master and from master we send to slave [So the master would know which is first]
-But lets say the master went down, then one of the slaves will becomes the masters, which will get the data, now what if in the middle the master comes up and gets the data. The master always knows it has older data.
I think only in Master-Master architecture, when you are using hashing for replication you need vector clocks during network skrewup.
Please correct me if i am wrong ?
