Greetings, dear readers of the blog! Today we will talk about the peculiarities of sharding on the GONT platform.
Scaling blockchain is the process of ensuring the growth of the blockchain system:
- From the point of view of infrastructure is growth of servers, so-called nodes in Blockchain network.
- In terms of the number of end users of the blockchain network. Accordingly, the number of transactions between users becomes larger.
Scaling the Blockchain network raises the problem of VM performance. With the increase in nodes and the number of operations of end users, the performance of the virtual machine is reduced.
As a result, there are 2 problems of blockchain network:
1. Against the background of increasing number of transactions is not enough capacity nodes to perform operations of end users, the queue for processing of information on operations of end users is formed.
2. Nodes is too much for the current number of end-user operations, so long as the transaction information goes through each node it will take a long time.
There is a problem of balance between security and performance.
Sharding is a technology of competent separation of transactional load on a network, it is a basic way of solution of these two difficulties – in terms of quantity of nodes and in terms of quantity of operations of end users.
The GONT platform allows you to use less nodes to process more user transactions per unit of time. This is due to the fact that not every node in the blockchain network should perform every task. Each nodes group performs the functions of one end process on a separate server, and this clustering makes transactions faster. Transactions like nodes can be clustered on the principle of ontological proximity, which will reduce the required resource needed to validate the transaction. For example, take the evolution of a regular computer. Outdated computers had only one central microprocessor. Over time, the architecture of the computer changed to separate mirochipy (a separate microchip-as a central processor, a separate microchip, believes audio, a separate microchip, believes video, a separate microchip, which Calculates 3d graphics, etc.) – This division into microchips is the process of sharding.
Computing capabilities of the computer, previously embedded in the central microprocessor, divided into separate microprocessors, “sharpened” on separate tasks. In general, computers have gained higher performance and the ability to perform different tasks in parallel. The next stage of separation is the formation of several cores in each microchip, for example, Intel server microprocessors, which allows to solve the problem of sharding within the microchips. The same system also works on the Gont platform. The standard blockchain virtual machine has a single-processor architecture in its classic form.
We solve the problem of sharding by introducing a large number of GVm nuclei within a single large blockchain virtual machine.
Each GVm core contains its own set of commands, instructions-this set of GVm cores differ from each other. Due to this distinction, each GVm core consumes its own amount of energy to execute its set of commands, instructions. To properly track the operation of individualGVm cores (i.e. Calculate the cost of running GVm Core) was introduced the concept of alternative service gas – ALGas (similar to the concept of energy).
Each GVm core requires its own volume of ALGas. Therefore Gont, unlike the old blockchain systems, contains different ALGas, depending on the required amount of energy required for the operation of each GVm nucleus. In other words, each GVm nucleus has its own ALGas. So, in unlike from the usual blockchain system, which has a centralized virtual machine. The machine has one microchip, which requires also one gas (on the example of a computer is the first computer with a single microprocessor).
GONT has a decentralized virtual machine, which requires different ALGas for each GVm core (for example Computer is a new modern computer with multi-core processors). ALGas allows you to accurately calculate the cost of each GVm nucleus.
Thank you for your attention! Have a productive Day!
Darya Studenkova