Three technologies for optimizing the data center network based on content

Three technologies for optimizing the data center network based on content

Network technology can be divided into multiple layers, from the physical link layer to the application layer. Different layers run different protocols and have different functions. With the development of data centers, network technology has been greatly improved, and networks have penetrated into every corner of the world. However, when we access the Internet, we often encounter network congestion and slow response speed, even a series of problems such as server error response or failure to access normally. To solve these problems, there are also many ways to start from the network. There are many content-based methods based on the application layer. The following describes three content-based optimization technologies in detail.

IDC

IDC-based content Introduction technology is one of the three major optimizations. The introduction of IDC content here refers to the direct introduction of Internet traffic through the data center, and then service optimization. For example, many large Internet sites will establish sub-centers in data centers across the country, so that local users can directly access the website, improving the access speed and user experience. In this way, the local data center needs to introduce the content of the IDC and directly introduce the traffic of the website locally. In this way, the data center needs to cooperate with major Internet websites, especially for websites with high access traffic, more traffic is introduced to the data center. The introduction of IDC content is the best way to fundamentally address the shortage of resources in the network. However, it is affected by factors such as the data center, bandwidth, operator, and CP negotiation, the requirement content source cannot be introduced in a short time. In addition, there are various problems such as insufficient depth of introduction of key websites, insufficient precision of introduction of hotspot websites, and unevenly distributed introduction of IDC resources, seriously affecting the access experience of users in the network. The introduction of IDC content is an extensive optimization that consumes a lot of network bandwidth and requires more storage servers to store data, increasing the burden on the data center, therefore, it is better to limit or filter the introduced content. Otherwise, it will put a great deal of operation cost pressure on the data center.

CDN

CDN is a new network content service system, which is created based on an IP network, content Delivery and services are provided based on the efficiency requirements, quality requirements, and content order of content access and applications. In general, CDN is website acceleration, which can solve problems such as slow Website access caused by cross-carrier, cross-region, low server load capacity, and low bandwidth. The main technical means for implementing CDN are high-speed cache and backup storage. The high-speed cache method is cost-effective and suitable for static content. The cache server can process static requests of most customers. The Mirror Site Server allows direct distribution of content, this method is suitable for static and quasi-Dynamic Data Synchronization. You need to purchase a new server and greatly increase the network bandwidth. Therefore, you cannot create too many backup storages. The most direct reason is the CDN optimization effect and the number of CDN nodes. The more nodes, the better the acceleration effect. Many CDN services are open-source, which provides fertile ground for the development of CDN, such as Baidu cloud acceleration, quickshield, 360 website guard, and jiasule. These are all free CDN services, it can be used in small and medium-sized data centers. For some large data centers, you still need to purchase professional CDN services. However, professional CDN services are expensive and can be used by big data centers. There are also a variety of CDN billing methods, including the lowest monthly consumption, the bandwidth, the number of requests, and the monthly or quarterly subscription, there are still some technical indicators that most people cannot understand and charge fees. In short, they are complicated. Most of the paybytraffic billing methods have time restrictions and will all be voided if they are not used up within the specified time. For small and medium-sized enterprises with poor traffic control, there is a considerable waste.

Cache

Cache stores a copy of the hotspot content accessed or downloaded by the user on the storage device. When the content is accessed or downloaded by the user again, you do not have to connect to the resident (source) website, instead, the cached copies are provided directly to accelerate access or download and reduce the bandwidth dependency between networks. Generally, Cache is divided into Web Cache and P2P Cache. Web Cache forces each request to be sent directly to the source server without passing the local Cache version verification, which is useful for applications that need to be verified, or use the reference of the latest data strictly. P2P Cache applies the traditional caching principles to P2P content. The basic idea is to Cache P2P content on the edge of the network, and filter out repeated P2P content using the cached content service for subsequent P2P requests. Deploying a cache device in a P2P network may obtain a byte hit rate of up to 90%. Using a P2P cache system can effectively reduce the usage of P2P traffic on the network bandwidth. At the same time, because the cache system is close to P2P users, user access does not have to go through the core network without quality of service guarantee, thus reducing user access response time and ensuring user access quality, improve user experience and attract users, so as to further protect the development of P2P technology. Of course, the Cache does not have no disadvantages. The Cache can provide normal services for static content, and it is prone to errors for dynamic content services, requiring refined control. The Cache has a certain lag in changes to Internet content, it will affect the normal service experience and cannot guarantee the service quality in real time. The Cache will reduce the willingness of the ICP to enter the data center and affect the introduction of the carrier's data center.

Content-based optimization technologies such as IDC, CDN, and Cache have different implementation principles and results, so they cannot be replaced by each other. IDC is more inclined to introduce and optimize large-scale traffic. CDN is content delivery, while Cache is refined management of content. If the traffic size is different, the optimization results will be different. Of course, each of the three optimization technologies has its own shortcomings. Therefore, each of them must have its own strengths. In a large data center, the three can coexist to make up for their respective shortcomings. In essence, Cache and CDN are both a turning and re-service system for Internet content resource access. The difference is that one is active, the other is passive, and the other is spof, one is a centralized Scheduling System Service. IDC needs to work effectively with Cache and CDN to optimize resource usage efficiency and traffic flow. By deploying three content optimization technologies, you can achieve unified and collaborative management of the entire network content source. First, the quality of service of the content is analyzed through the distribution Dial Test of the whole network content source. Based on the introduction of IDCs, CDN distribution, multidimensional analysis, and collaboration principles, the Cache can be cached on demand, and the Cache/CDN content can be mutually exclusive, optimizing the full-network content source service. Three technologies work together to greatly improve the data center optimization effect.