Ripple recently announced the new feature of History Sharding for their version 0.90.0. The new feature reportedly allows storing historical data for individual Ripple servers instead of storing entire history. The company claims that due to the new feature, data storage needs will be reduced for the mentioned servers. Currently the saved space is reaching up to 8 terabytes.
Meanwhile the official message also confirmed that the new feature of sharding was not added to replace a ledger store. Instead this is reportedly “a step towards scaling the blockchain and achieving better results…”
The original message also stated: “the history sharding feature, enabled in rippled version 0.90.0, addresses this issue by distributing history into segments called shards. A shard contains all of the data for a range of ledgers. Using the history sharding feature, individual rippled servers can contribute to storing historical data without needing to store the entire history”.
In addition to providing details on the new feature, the official message also explained the actual problem that the new addition is there to solve. As explained, due to the fact that more users start adopting the new technology of blockchain, the network size keeps growing. And the servers, that maintain this growing network on XRP Ledger, store over 8 terabytes of data in order to store its full history. So the key issue is that each server has to have huge amount of resources to store the entire history. This may soon bring to list of issues connected to the efficiency of the blockchain.
So How Does the New Feature Work?
The company introduction suggests that “instead of replacing the ledger store, sharding creates path connecting directly to XRP Ledger Network. Hence, the history of the entire network remains intact while servers only agree to maintain shards. The servers confirm that they have all the data they agreed to maintain and produce proof trees or ledger deltas”.
The message also confirmed that shards are selected randomly. Meanwhile a normal distribution curve stores the full history of the ledgers. As a result the network is able to maintain the history.