What are the characteristics of data center business data?

The data center carries a variety of business types, such as: payment, monitoring, management, website, data analysis, etc., the data center is almost omnipotent. Even though these business forms are very different, but essentially are data, which is the core characteristics of the data center, the reason why the data center is called data center, is because it is the center of the data processing, data center, all systems are centered around a variety of data to start, there is a special study of the data technology. All application business data in the data center can be divided into three categories: voice, video and data. Here are the characteristics of these three types of data in detail, according to the characteristics of these data can be effectively adjusted to the data center network so that it can better serve these data.

1. Voice data

Voice data is extremely sensitive to drop and delay, if the amount of drop is relatively large, the voice will sound intermittent, can not hear clearly, more noise, if the network delay is relatively large, the voice will produce echo. So the data center to carry voice data services, you need to meet certain requirements in terms of delay and discard rate.

2. Video data

Video data has been growing rapidly in the past few years, almost exceeding the sum of other data traffic. Video data is large, rich in content, and intuitive to display, so many applications like to use video data. Video data has a certain degree of suddenness, non-benign preference for adventurous, equally sensitive to discard and delay, like a lot of high-definition video, in order to ensure the clarity of the need for a very low discard rate.

3. Common data

There are so many different kinds of data applications, different applications have different traffic characteristics, different versions of the same application may have huge differences in traffic characteristics. This type of data is characterized by both smoothing and bursting, and both compliance and adventurousness, although generally this type of data is insensitive to latency and discarding.