At present, several well-known live broadcast websites in China have monthly activities of more than 87 million. If the monthly activities of tens of millions and online activities of millions are calculated, this will mean the consumption of 1.5T bandwidth, and nearly 30 million cash will be spent on bandwidth every month. As the business volume continues to increase, the cost will only increase. At the same time, there is a more serious problem. The competition in the CDN field is becoming more and more fierce. When users understand the correlation between cost and service quality with professional eyes, how can they choose the most suitable manufacturer among many CDN operation service providers?
Intelligent scheduling Ting Yun Controller is only born to save costs.
▲ Wu Jingtao, Vice President of Ting Yun Technology, released Ting Yun’s new product-Ting Yun Controller.
In order to solve this complicated situation, Ting Yun’s brand-new product, Ting Yun Controller, came into being. After in-depth analysis of the big data collected by APM in Ting Yun, it achieved intelligent scheduling of resources and provided users with real-time low-cost and efficient resource allocation suggestions. The first phase product of Ting Yun Controller opened the CDN scheduling system in Ting Yun-Ting Yun Controller for CDN, and used Tingyun big data to conduct credit reporting for CDN service nodes, so as to schedule CDN in real time with high efficiency and low price.
▲ Real-time CDN scheduling
Billions of big data support is only for real-time control of CDN.
At present, Ting Yun can collect 20 billion pieces of mobile performance data, 300 million pieces of browser application data and 30w+ national monitoring nodes every day, which together form the basic data of CDN and CDN performance monitoring in Ting Yun. According to these real-time CDN node success rate, user throughput, actual access time, price and other data, we can provide a good evaluation and use possibility for CDN node selection.
So what will happen after using these data to make intelligent scheduling? Let’s take an example. When we choose a picture of a real user and two CDN operators to do normal stress tests, what will happen to the real performance after using Ting Yun’s CDN scheduling?
Performance average comparison: the download time of real pictures on the client side is reduced from the average of 350 milliseconds to 191 milliseconds, and the performance is improved by 40%.
Average success rate comparison: Compared with the success rate before and after scheduling, once the service node has availability problems, it can immediately help to switch the traffic to other nodes, and we can see that the success rate is greatly improved.
Comparison of the number of nodes: the change of service nodes. From 264 nodes of two CDN service operators to 500 nodes of five operators, the available nodes become more. But in fact, the scheduling does not need complicated actions, and the Ting Yun Controller system can be fully automated.
Actual cost comparison: actual cost change. The introduction of new CDN operators may bring the possibility of cost reduction, but with it comes the risk of availability. Using Ting Yun CDN scheduling system can greatly reduce the cost while ensuring availability and performance.
We can see that after using Ting Yun Controller for CDN to intelligently schedule the usage of CDN, under the condition of ensuring availability and performance, the cost is reduced by 40% and the efficiency is improved by 30%, so that your cost will not increase with the increase of business, and the proportion of cost increase will be reduced. This is the concept and value of Ting Yun Controller, which takes the real user experience as the core, ensures the performance, availability and cost of every node running on CDN, generates a credit system, and helps users to make real-time decisions based on service quality and actual value when selecting service nodes. In the future, the product line of Ting Yun Controller will be extended to Cloud and active/passive DC user intelligent routing to build a truly intelligent second-pole user experience era.