IPFS is essentially a Distributed Hash Table (DHT).
DHT
Used to record and describeCIDs
And the user who uploaded the data (IP
).- Because in the
IPFS
There is no storage central node in the network, soDHT
It’s distributed storage. - As information is being
IPFS
Like distributed storage on all nodes in a network,IPFS
Every node in the world will store itDHT
A subset of.
When we talk about a user uploading information to an IPFS network, what it really means is:
A CID that tells the IPFS network that a piece of information has been added to the DHT;
A mapping relationship is added in the DHT: CID of the new information and the IP address of the user.
If a user needs to download data, locate the IP address in the DHT based on the CID(content ID) and download the data.
The speed and reliability advantages of IPFS networks are due to the fact that different users can upload the same data, which means that different IP address mappings of the same data can exist in DHT.
Thus:
IPFS
Network data can be downloaded nearby, which will undoubtedly greatly improve the speed of data download.- Even in the
DHT
If the user machine with data mapping relationship is in the shutdown state, the data will not be available, even if a node user decides to give up local storageDHT
If the CID corresponding to the CID recorded in, another node user can immediately replace the node user to store data. This undoubtedly improves the reliability of data acquisition.