Most P2P blockchain networks replicate all the data (transaction data, algorithms, software, etc.) in all the nodes of the network. Bitcoin, Ethereum, and many more have achieved this by sacrificing scalability & speed.
However, today there are billions of transactions happening every second around the world (payments, IoT, supply chain, etc.). Given the material and physical constraints, it is not possible to store all these transactions in all the nodes of the network.
Hence, many blockchain networks today instead of storing the data in all the nodes, store in a few selected nodes. This increases the scalability of the network. This is called by various names like sharding, sidechains, plasma, …
Note: The underlying algorithm to select a few nodes is different and depends on the consensus mechanisms.
A blockchain network is a form of database with a pool of nodes where each node represents an individual data server. When a blockchain is sharded, this means that the blockchain is divided or broken into small blockchains or shards. Where each shard would have unique transactions, smart contracts, data stored in them. The nodes will be assigned to individual shards to verify and store transactions, instead of all the nodes verifying all transactions on the network.
Sharding => Break up the blockchain into small segments called shards until each shard is more managable.
Sharding will lead to high TPS and hence overcome the scalability issues.
Yes, sharding is good but what is the problem?
Data availability and data consistency of sharded blockchains are less than traditional blockchains. For example,
- Due to P2P network problems like internet connection problems, electricity cuts, many more... The nodes go online and offline all the time.
- When a shard in the blockchain network goes offline, the data belonging to that shard is lost forever.
Hence the security of the blockchain is compromised. Contrary to this the network might not recover from the damage caused by a few nodes.
Uniris, the worlds first self-healing blockchain
Due to the reasons above, blockchains should be able to self-heal no matter what the network conditions are. Hence, Uniris introduces a self-repairing mechanism!
The mechanism is simple, let us compare normal sharding with Uniris sharding
From the fig, if a shard in Normal sharding goes offline then all the data belonging to that shard is lost forever. But in Uniris sharding, if a shard goes offline the data is still available and the network continues to run. Once the nodes of the shard are back online then they will update the view and store the data that they are supposed to store.
- To preserve an unlimited network capacity, a very small amount of information is replicated on all nodes.
- To ensure automatic repair, nodes use the same information, formulas and algorithms.
- This mechanism allows any node in a few milliseconds to self-repair.
To have an in-depth understanding of Uniris's Self-Repairing mechanism please refer to the yellow paper.
Uniris is not only a pioneer in self-healing, but also in (12 international patents)
- Scalability => Capability to handle UNLIMITED transactions per second
- Traceability => Zero knowledge proof
- Decentralized => Randomness ensures complete decentralization
- Sustainable and Green Solution => Energy consumption is 3.6 billion times less energy consumption than Bitcoin
- Physical and Cyber Access Management => Uniris blockchain is coupled with 'Public Biometric Identification' (Patented)
- DApps Compatibility => Market Place, Identity Management, etc.