in web-based project,how to design databases to make it possible to handle of millions of requests.
firstly, in MySQL databases ,we consider the website usually more query requests and less writing requests relatively .we can divide these two operations into two systems ,read database and write database. in application ,we can make the requests access different databases
secondly ,use the MySQL feature "replication" mechanism rapidly to copy data from the main database to the secondary databases., one main databases can be more secondary databases , it is the real time synchronize between the main database and secondary database
thirdly , there can be many write databases which can be shared with the applications ,it can solve the performance bottlenecks and single point of failure
fourth ,there can be many read databases ,with the load balance cluster mechanism ,to make it high available ,reliable ,better performance