On the evening of July 13, Beijing time, the website 404 of “blackout” at Station B was unable to open, and the news quickly became refresh, ranking first on the list of most popular searches.



Acfun and Douban also crashed, so it seems that the sites that crashed were heavy traffic sites, losing thousands of dollars a minute.

The official media of Station B said that part of the server room of Station B received a malfunction, which made it impossible to access. The technical team immediately started troubleshooting and repairing, and the server has now returned to normal operation.

This news out of the B station US stock gains immediately narrowed!

It seems that the underlying protocol of HTTP network that we are using now is not reliable and can not withstand twists and turns and tests. The underlying network protocol we now use is called HTTP, which is a centralized storage mode, which means that all the data are stored on a huge storage device. Once there is any problem with this huge device, data loss or data anomaly will occur, just like the phenomenon of B station and Douban. When surfing the Internet, we often encounter 403, 403 and so on. In fact, this kind of situation is the loss or corruption of data. It can be very expensive to lose a very important file stored on HTTP, because once the file is lost or corrupted, it is essentially unrecoverable, and HTTP has many disadvantages besides being easy to lose. For example, the central server is hacked, all the data will have no privacy at all, from personal privacy to national secrets will be stolen. Your secret photos uploaded to the cloud could be shown to 7 billion people around the world. Now we’re spending billions or more on cybersecurity.

Network speed is slow and storage cost expensive are the disadvantages of HTTP, now a few G cloud space rent just big hundreds of dollars a year, and that’s the disadvantages of centralized storage, we browse to browse a web page at the same time the number too much will cause the loading speed is slow, and this is not a network speed problem. Big brother for example, HPPT is a road, on this road will be more traffic jam



Compared with IPFS there is no such problem, the above IPFS is a decentralized block chain technology of distributed storage agreement, it is the concept of a data are stored in various places, this directly to ensure the safety of data, if the invasion is only a small storage data, IPFS solved the problem of the HTTP excessive redundant space now. So if you want to watch a movie on IPFS, IPFS will ask the whole network, “Does anyone have a hash for this file?” There will be nodes on the IPFS network that return the file, and you will be able to access it. In plain English, this means that the same file only needs to be stored once on the network, and then stored and published in the future for all IPFS users to use as long as they want.

Since IPFS is distributed storage, it does not have the sudden failure or slow loading of HTTP. For example, HTTP is one road, and IPFS is an infinite number of roads. If one road is blocked, the other will be followed. In the process of data storage and exchange, IPFS is undoubtedly a major revolution of blockchain. As a contributing and valuable network, what has overturned is giants such as Amazon Cloud. It belongs to the data-contributing type. Because they can really share data on a large scale.