The Ethereum blockchain network has been overpopulated in recent months and that can be regarded as both good and a bad sign. The positives are that it just proves its utility and the negatives from the overpopulation are the transaction backlogs and higher fees.
The Ethereum blockchain network has been overpopulated in recent months and that can be regarded as both good and a bad sign. The positives are that it just proves its utility and the negatives from the overpopulation are the transaction backlogs and higher fees. But there is something else that bothers developers – the growing number of users leads to huge storage costs on those who want to run a full node.
All computations that should be memorized by the computers enjoying the platform are held in a database, called the ethereum state. Since storage costs increase along with the number of computations, fewer people choose to run a full node. Accordingly, this may disrupt the decentralization of the network.
Luckily, ethereum engineers are familiar with the problems and they do approach them. Protocol-level fixes like sharding aim at shrinking the database. Other stakeholders who run the software that enables users to interact with the blockchain need to find a solution for the state of the database as well.
Afri Shoedon, a technical communication manager at Parity, admits that growth is happening so fast that by the end of the year small devices might not be able to run the platform. Geth and Parity are the largest Ethereum software providers and they have both been trying to cope with the database issues by releasing updates recently.
Parity’s update reduces storage requirements. When the software memorizes ethereum history it produces temporary files and the update eliminates the unnecessary temporary files. This leads to faster synchronization and on the other side, the software is manageable on hard drives instead of solid state drives. Almost two years after the last time it was possible to run Ethereum on a hard drive due to storage requirements now this problem is solved.
Meanwhile, the independent developer Alexey Akhunov is rewriting the Geth client code. His goal is to ditch the unnecessary repetitions clients do when processing the overall state. Furthermore, Akhunov proposes other solutions as well. He states that it is possible to rewrite the software so that it can use random access memory (RAM) and thus become faster. As you may have guessed, Geth’s in-house development team is also optimizing the platform in hopes to make friendlier and faster.
Ethereum engineers are also working on the so-called “stateless clients”. Those types of clients will store only a fraction of the overall state instead of the whole state. Both Buterin and Szilagyi support this concept because it will ease the pressure on individual clients by splitting data into smaller pieces, stored in various nodes.