Showing entries 1 to 1
Displaying posts with tag: fault-tolerant systems (reset)
Scaling out and creating fault tolerant systems with MySQL replication

A few days back, reading the article on the Moore's law, I thought it is not too difficult to extend this for an analogous situation, e.g. the amount of data being cached by a search engine like Google also grows exponentially. With changing times we need to have solution that would adapt to the changing data requirements and availability solutions like virtual infinite scale-out like cloud. Even though virtualization is not exactly a database concept, it's  not long when we will have it as a standard for RDBMS. This article is however  about using MySQL replication to scale out reads and make it fault-tolerant.

MySQL replication is essentially asynchronous  i.e. it doesnot guarantee zero latency between the master and the slave(s).  This, by extension means that in case of a master crash, data may be lost if we decide to promote one of the slave to be the master and that is why while scaling …

[Read more]
Showing entries 1 to 1