The initial stage of the site is small, only a low profile server can be configured, applications, databases, files and other resources are on a server. With the development of the site, you can always adjust the configuration and the number of ECS, no need to worry about the lack of resources in the server when the business highlights the problem.

The cloud server ECS and object storage OSS collocation, OSS as a static picture, video, download package storage, in order to reduce the storage cost, at the same time with CDN and load balancing, can greatly reduce the user access latency and reduce bandwidth costs, improve availability.

Using a high profile I/O optimized ECS server, while using SSD cloud disk to support high I/O concurrency and higher data reliability. You can also use a number of slightly lower I/O optimized ECS server, load balancing, to achieve high availability architecture.

The difficulty lies in the deployment during peak traffic, but also to control costs. Through elastic expansion, the ECS instance is automatically added when the business grows, and the ECS instance is automatically reduced when the service is down.

E-MapReduce is built on the ECS, such as log analysis, data warehouse, business intelligence, machine learning, scientific simulation and other business needs. You can use E-MapReduce to run the commonly used computing framework (for example: Hadoop, Spark) to interact with other data storage services such as OSS, RDS in the cloud.