The Filecoin mainnet went live, bringing the entire distributed storage sector to a high-light position. As an infrastructure in the era of Web3.0 and the digital economy, the storage market using data elements has brought new business models and economic formats, but distributed storage can only be regarded as an important starting point, and the real value generation lies in it The next step.
The premise of data becoming an asset is the growth of quantity and necessity
As the development of 5G, AI, and the Internet of Things promotes the era of the Internet of Everything, everything needs to be recorded and expressed in data, and data will change from a single internal small data to multiple dynamic big data. Especially with the industrial Internet 4.0 and the access of new industrial terminal devices such as autonomous driving, smart cities, and smart homes, the data will show explosive growth.
As for the amount of data, Intel has predicted that the total global data will reach 44ZB=44 trillion GB this year. In addition to the IDC forecast, the scale of the global data circle will increase from 33ZB in 2018 to 2025. 175ZB also pointed out that unstructured data such as text, pictures, and video will have a higher growth rate, and its proportion in the overall data circle will continue to increase. This also means that a more advanced Internet infrastructure is needed to collect, store, and utilize data.
The success of distributed storage in narrative lies in its initial realization of the idea that data producers and owners control data with the help of a decentralized business model, and the use of incentive models to mobilize a potentially very large and active data storage market. But the problem is that the data itself has no value. Only after processing, calculation, and analysis can it generate value and be given a price. So we can’t just think about where the data is safer, the ultimate goal should be how to use it after it is stored in order to tap the value of the data under distributed storage.
The optimal solution for data calculation is always changing
In fact, people’s pursuit of computing power growth has never stopped from beginning to end, especially after the emergence of the Internet, this computing power battle has become more and more fierce, and computing power and processing have alternated between a centralized architecture and a distributed architecture. With the large-scale commercial use of the Internet, corporate IT, and smart phones, a wave of cloud computing based on large-scale centralized server groups has been ignited. This process includes from the early manual code compilation and calculation, to the functional microservices of module scripts, to the emergence of specialized IDC data centers in response to high-concurrency data calculations, and later in order to further release the long-term value of resources. The machine is configured as a service, which has become the prototype of the cloud computing we often mention now. The advent of cloud computing has brought unexpected benefits. You don’t need to install various software on your computer. You only need to access the “cloud”. Internet giants can provide ready-made computing capabilities. Associated, cloud storage services are also very popular.
Everything has its pros and cons. Cloud computing and storage have their own limitations in the development: First, the industry is developing monopolized by Internet giants, and it is difficult to mobilize a super-large-scale resource market, which includes storage resources and computing power. And so on, secondly, the centralized storage method, not only the stored data cannot be open and transparent, but also easy to receive malicious attacks, and the safety factor is low; again, the cloud computing paradigm of data collection-cloud center-data consumption is inevitable. High latency, unstable network and low bandwidth issues. In the case of a surge in data volume and exchange speed, it will cause inefficient results.
Based on the continuous development of technologies such as the Internet of Things and the continuous increase of data, the industry is always looking for the best solution, expanding from clustered to decentralized paths, such as distributed computing frameworks, virtualization Computing, etc., to a certain extent, overcome the drawbacks of centralized computing, and new requirements (such as the Internet of Things and enterprises) and the potential of ultra-low latency services in the 5G era are bringing computing closer to end users, which means that the business model is shifting from network-based The shift from the center to the workload-centric signifies that the distributed transformation of computing power and resources close to the deployment of customers will become a trend, and emerging edge computing will take the lead.
Can edge computing become the best practice for distributed storage
What is edge computing, Gartner defines edge computing as “a part of distributed computing topology, where information processing is located near the edge, that is, where things and people generate or use this information.” To put it bluntly, it is closer to data A way of calculating the source. For example, cameras with visual processing functions, wearable medical devices that send data to mobile phones via Bluetooth, etc. all use edge computing.
Compared with cloud computing, edge computing is closer to the terminal, reducing transmission costs, while decentralizing broadband and reducing latency. In addition, edge computing provides more communication paths than centralized mode. This distribution can better ensure the flexibility of data communication. The most important thing is that the edge computing platform provides a new entrance, which can make the smart terminal in the idle state work and realize the sharing of physical resources. With many excellent features, edge computing will become one of the main growth engines for the server and storage market in the next decade and beyond.
So why does edge computing need distributed storage? First, the Internet of Things, 5G, AI and other technologies have brought more application scenarios and smart terminals and faster connection speeds. In this context, various terminals will access the Internet and serve as edge computing computing facilities, providing high-speed and With the timely calculation of foundations, more data will be produced. How to solve the problem of large-scale data storage will be a problem that must be faced. Existing IoT storage devices mainly rely on local storage units to store data, and their storage capacity is limited and storage expansion is difficult.
The distributed storage project can just solve this problem. When edge computing generates data storage requirements, distributed storage can rely on global access nodes to provide nearby storage services, and provide the best and closest storage resources on demand and quantity. The edge storage solutions it provides can integrate different storage services. The vertical expansion becomes horizontal expansion, which reduces bandwidth and energy consumption and improves storage response speed, so as to optimize storage on demand at close range. Projects like Filecoin and Storj are solving such problems. It also improves the utilization of idle resources to a certain extent, and stimulates users’ enthusiasm to make full use of data to create value.
The cooperation of the two is complementary. Specifically, edge computing can effectively mobilize idle broadband and computing resources, and the incentive system of distributed storage solutions can mobilize existing social stock resources as the basis of edge computing. Facilities, which will greatly shorten the construction period of edge computing, form an organizational network, and solve the problem of effective resource allocation. This is of great significance to edge computing in its infancy. Moreover, the mesh structure of edge computing and the mesh structure of distributed storage can form a good fit. Each computing terminal can obtain the space of a lower cost storage node nearby. If distributed storage can solve the current data encapsulation, Read speed and privacy protection issues, then it can reduce the bandwidth and storage resource occupation of the backbone network, and the cost is also lower.
This advantage is very suitable for the construction of future smart cities. Just imagine that the terminals and collaboration relationships involved in the IoT network formed by a smart city are extremely large and complex. In the process of construction and rapid development, only the government It is difficult to build a well-formed infrastructure environment for this network from the beginning. Moreover, the risk of a centralized solution to a super-large network such as a smart city is extremely high. A major problem may cause the entire city’s system to be paralyzed. Here it can better reflect the necessity of distributed storage, one is the cost and the other is the risk resistance brought by the distributed system.
More than that, after solving the basic needs, the two can also create more application cases. First, the personal storage market and small data centers will become industry trends. Data show that in 2020, the number of Internet users in China will be more than 900 million, and the number of mobile Internet users will reach 1.3 billion. The monthly active users of personal online disks exceed 100 million. The personal storage market still has a huge potential user base and available storage space. Faced with the rapid growth of data scale, edge cloud computing uses distributed file systems to put personal idle storage resources into the network for market exchange, which will become one of the first areas for distributed storage to enter. In addition, the innovative imagination of the combination of the two may occur in every item in the home, as large as the business scene of the smart city, and lay a solid foundation for the service system that may be born in the future. In short, the second half is good The drama is worth looking forward to.
