Challenges of decentralized storage

Centralized storage is the mainstream way of data storage nowadays , Even cloud storage, known as distributed storage , In essence , It still belongs to centralized storage . This storage method that relies on storage service providers to provide servers has gradually failed to meet the growing amount of data and people's demand for data storage .

The decentralized network built by blockchain has become a new exploration direction , Decentralized storage, which relies on massive nodes to provide storage space, has brought a breakthrough for data storage . But this change was not achieved overnight , The implementation of decentralized storage still needs to face many challenges .

Data storage format

There are two main forms of storing data in decentralized networks , One is to store complete files directly , The other is to use erasure codes to store fragments of files . In terms of technical difficulty , The former is easier to implement and complete , But corresponding , There are also more problems .

The nodes in the blockchain network are jointly formed by users , Although all nodes in the overall network have high security , However, a single node is insufficient in both storage performance and security protection . There are great hidden dangers in storing complete files directly on a single node . therefore , More projects use fragmented storage , For example, today's well-known decentralized storage Filecoin、Swarm All use erasure codes . In this way of storage , Don't decompose the data into fragments of constant size , Scattered in different nodes , And each segment will be encoded with redundant data , Further enhanced the security protection of data .

Store trace

Decentralized network nodes are scattered in different regions , This makes it need stronger tracking ability for the data stored in it , In order to better retrieve the data 、 Inquire about .

On storage tracking , There are also many different schemes . such as Filecoin Use local blockchain to track stored orders , At the same time, the storage node is also used to maintain the local network location list ;Swarm It writes the data address directly into the data block , During data retrieval , The network can quickly search according to the data .

Proof of storage

Decentralized storage uses user nodes to provide storage space , Users can get paid by providing nodes , But at the same time , In order to prevent users from leaving at will after accepting the stored order , Nodes are also required to submit valid storage certificates within a fixed time , To verify the validity of the storage .

Storage proof is a mechanism used to ensure long-term storage of data in decentralized storage . Different storage networks also adopt different ways of proving .

Filecoin Using a method called duplicate proof , This is a method of sealing data in a closed storage sector . And get a duplicate fragment in the sealing process , Through this copy fragment and the original data, it is mutually proved that they are unique , To verify that it is stored on the network .

and Swarm The data block is stored in Merkle In the tree . When storing data , The root hash of the tree proves that the file will be correctly partitioned and stored .