Nion Network — The Genesis
Ever since the invention of Bitcoin, the shift towards decentralization and trustless systems has been gaining momentum. Granted, much of the progress is covered in layers of intricate relationship between innovation and investment. However, pealing the layers of the onion reveals it’s about much more then that. It’s a movement towards a decentralized world in which the moto “Trust but verify” echoes strong in every system.
A common question on the path to understanding the Bitcoin protocol is
“Why are miners computing something so useless?”, and analogue “Can we make a system, where miners compute useful things?”. These are reasonable questions, and recently, with the growing concerns about the environmental impact of mining, very valid questions.
Another issue is, miners are all solving the same problem. Ethereum’s smart contracts can be considered an improvement. However, despite being touring complete programs, they are quite inefficient as all miners have to execute each contract to verify the state transition.
The scrutiny about mining energy, and hardware inefficiencies calls for new, more efficient consensus mechanisms such as Proof of Stake(PoS). Assuming the security of PoS, and PoW is comparable (a debate for a different venue), the consensus layer will be much more efficient. However, the smart contract executions will still require all validators to execute each contract.
Another good question would be “Can we still have all the nice features of blockchain protocols, and get rid of redundancy in computation?”. At Nion Network, we begun work on the idea in 2017. It was purely for academic purposes, and to this day this changed only slightly. We envisioned a decentralized system with a vote based consensus mechanism as a platform for applications with no computational redundancy.
In more plain English, a decentralized cloud computing platform.
Our solution was quite simple really. We aimed to address the following key problems:
- Address the problem of language heterogeneity
- Optimize resource allocation with decentralized and autonomous migration/orchestration
- Provide verifiability through transparency
The result was a permissioned blockchain with a federated vote-based consensus mechanism. The nodes were not only responsible for contributing to the security, and participating in the consensus. They were also responsible for sharing their resources by running containerized applications. In order to have your application run, you only had to build it as a docker image, and ask any node to run a containerized instance regardless of the language or tech stack your application was written in.
On each new block, the validators decided on a plan to migrate containers where needed to improve resource allocations across the system. Due to the deterministic nature of the migration algorithm, the input, and outputs are written into a block, and can be verified providing a history of the entire lifecycle of each application (container). However, there were a few assumptions made that, which is quite common in academia.
- No byzantine behaviour of nodes
- Applications are stateless, and can be migrated at any time
- Applications do not assume any persistency in network identify
In a permissioned system there was need to have the protocol tolerate byzantine faults. We just assumed nodes would be honest, and would not drop before, or during a container migration. Containers were migrated very quickly as there was no need to transfer the entire state, and data, which is only useful for some application types like microservices. And there was no way for the container to have a reliable public IP address or any fixed networking parameters as the host kept changing.
Nion Network, is the next evolution of the protocol. We are moving towards a permission-less network, where the consensus will be secured with PoS to address byzantine behaviours. Migrations will allow transferring the state, and on deployment, fixed network parameters can be requested. We are also moving from an academic only project, to a fully production ready, opensource platform for the next generation of cloud based applications.
We are working on a technical explanation of the protocol, which will be published after the pending peer review journal submission. Stay tuned!
Also, you can check out the public dashboard we use to monitor our private testnet deployments at:
https://dashboard.nion.network/grafana
Follow the project progress on:
Or join the discussion on
Telegram: https://t.me/NionNetwork
Twitter: https://twitter.com/NetworkNion
