Developer Update: Lition Blockchain — March 2020
One of the most noteworthy action points on our 2020 roadmap is fixing the IBFT consensus issue which led to the mainnet being downgraded to beta status. During Q1 2020, the development team worked hard to stabilize our blockchain and close the transaction batches. In this Developer Update, we will share with you in detail, our efforts so far at re-engineering our blockchain back to full production Mainnet status.
To ensure a permanent flow of transactions, we implemented a new architecture for the Lition network, decoupled from the blockchain, in order to implement its tasks independently. This new implementation includes validations and alerts to guarantee a high level of availability.
In parallel, we have taken a lot of actions to solve the consensus issue that causes the blockchain to stop producing blocks after running for a while. You can find more technical details about the Consensus issue here.
Progress until now
The first thing we did as a team was to carry out extensive theoretical research on different implementations of IBFT consensus in blockchain to identify those that we believe we can build on and take further.
The next step was to analyze other consensus mechanisms that we could potentially use as a reboot if we ultimately are unable to find a viable solution to the consensus issue of the current one. An example of consensus mechanisms that we have looked into is Raft which is showing some promise. We are currently validating its security as it relates to the high requirements of our enterprise use cases. Casper FFG, or more popularly called Vitalik’s Casper the Friendly Finality Gadget which is a hybrid POW/POS consensus mechanism is another one that we’ve researched in detail.
In order to avoid recency bias from our current blockchain, we started a brand new testnet for the network which we are using to test all the different fixes developed for the blockchain. This enabled us to gain unbiased results that our development team has been comparing and improving upon to determine the best approach to concentrate the maximum efforts on and which to abandon. The testnet we used is modeled after our current blockchain with a creator, 3 validator nodes and one transactor node. You can find an instance of the testnet here.
Collaboration with Quorum
We have been working closely with the Quorum community to analyze several theoretical solutions that have been proposed for the consensus issue. One such solution is the analysis of the correctness of Istanbul BFT explained further in a paper published here. The Quorum community has been extremely helpful and have shared their discovery logs with us, providing extensive materials to work on together. The mutual collaboration has also led to the identification of other issues which we will further explore.
We are however seeing results from this cooperation as we are now working on a significant upgrade of our version of Quorum based on their most recent release. This new release enables new, powerful features such as importing and exporting chain data that helps to provide more information about blockchain issues and also analysis of fixes. It also enables Quorum plugins that also improves our capacity to develop solutions on the platform.
Final words
We have come a long way in understanding a notoriously complex and hypothetical problem with the help of the originating community and have made significant progress in testing several capable and promising measures to fix this problem. There is enough progress to give us confidence that we will be able to achieve full Mainnet stability in a matter of time.
