Shulou(Shulou.com)05/31 Report--

This article mainly introduces why Facebook chose Hadoop and HBase, the article is very detailed, has a certain reference value, interested friends must read it!

Strong horizontal expansibility. For massive data storage scenarios like Facebook, capacity expansion is almost common, and HBase can make data expansion very easy.

High write throughput is supported. Facebook has a large number of messages and a large number of writes every day.

Strong consistency can be guaranteed in the same data center. Facebook uses HBase to store message data, and business needs a consistent data store (which is one of the reasons why Facebook does not adopt Cassandra ["WefoundCassandra'seventualconsistencymodeltobeadifficultpatterntoreconcileforournewMessagesinfrastructure" fromquora])

It has good random reading performance. The business logic of the messaging system results in a lot of random read operations that penetrate the cache layer.

High availability, fault recoverability. Due to the large amount of data, there may be many distributed machines, failures or some daily upgrades will be more frequent, and a system with high availability is needed.

Error isolation. The error of one node will not affect other nodes, and the disk failure will only affect the corresponding small-scale data.

Provides atomic read-modify-write operations. Atomic increment or modified operation after comparison is very convenient for many business processes.

Provides the ability to get a range of data. For example, functions such as getting someone's last 100 messages are a common requirement in messaging systems.

Of course, the following areas that Hadoop and HBase are not very good at are worth mentioning:

Disaster tolerance under the separation of the same data center network. There is a problem with the network in the same data center, which makes it impossible for nodes to communicate properly, which can usually be avoided by configuring some backup network equipment.

A data center failure does not affect services. This situation is even more rare.

Real-time data exchange between multiple data centers. This is not very realistic, which is usually achieved by using the Cache layer to achieve real-time access to senseless data.

The above is all the content of the article "Why Facebook chose Hadoop and HBase". Thank you for reading! Hope to share the content to help you, more related knowledge, welcome to follow the industry information channel!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.