Sharding Pinterest: How we scaled our MySQL fleet

https://medium.com/@Pinterest_Engineering/sharding-pinterest-how-we-scaled-our-mysql-fleet-3f341e96ca6f


Comments

Post a Comment

Popular posts from this blog

What are Master and Slave databases and how does pairing them make web apps faster?

What are Master and Slave databases and how does pairing them make web apps faster? A database is "slaved" to a "master" when it receives a stream of updates from the master in near real-time, functioning as a copy. The "slave" must simply apply the changes that the master validated and approved. In principle, one could create a master-slave setup by transferring files really fast from one server to the other; but in practice, each database has its own specialized replication protocol. There are many reasons a replica makes queries return faster. One is that the master's CPU is less burdened by queries while the replica's CPU is less burdened by writes; so there is more CPU available to do work. Another is that data, while nominally stored to disk, is cached in RAM when that is possible; and indeed the OS does this whether the database asks for it or not and can be unpredictable about it. On the master, recently written portions of the database ...
Read more

Storage Scalability ->Vertical scaling and Horizontal scaling Sharding

Storage Scalability We try to explain some of the terminologies in simple words. Lookup wiki for a more formal definition. Replication  : Replication refers to frequently copying the data across multiple machines. Post replication, multiple copies of the data exists across machines. This might help in case one or more of the machines die due to some failure. Consistency : Assuming you have a storage system which has more than one machine, consistency implies that the data is same across the cluster, so you can read or write to/from any node and get the same data. Eventual consistency : Exactly what the name suggests. In a cluster, if multiple machines store the same data, an eventual consistent model implies that all machines will have the same data eventually. Its possible that at a given instance, those machines have different versions of the same data ( temporarily inconsistent ) but they will eventually reach a state where they have the same data. ...
Read more