What is read/write split? How is it important for scalability?
Most users scale MySQL or SQL Server by adding more secondary databases. Doing so in a conventional setup requires users to modify their application to redirect and load balance read queries (the SQL “select” command) between both masters and slaves and to ensure that any write queries (SQL commands “insert,” “update,” or schema changes, for example) are sent to masters. This process is complicated, time consuming, and error prone. ScaleArc enables the capabilities in its software rather than forcing you to program them into the application layer. With ScaleArc, you configure specific servers in a cluster to act as either a master or a slave, and ScaleArc automatically routes read queries to both masters and slaves while making sure that any write queries go only to masters. This process is entirely transparent to the app and database, so it does not require any app changes. Transactions are treated as writes even if all of the SQL calls within a transaction are reads.