From Start-up to Scale-Up: How Tokeny Prepared for Large-Scale Tokenization

Scalability is one of our top priorities as we are working with large financial institutions with very high volumes anticipated. We regularly run intense scalability tests and a few months ago, we concluded that we should reduce our dependence on some third parties to provide faster and enterprise-grade solutions. This newsletter explains how we became more autonomous and scalable by running our own blockchain Indexer and nodes:

Blockchain data Indexer

In order to build the real-time cap table of token holders with fully accurate information, our T-REX Platform needs to scan and read the blockchain to index relevant transactions happening to the ERC-3643 tokens and ONCHAINIDs. Technically, it can be very challenging for complex operations, and tools like The Graph, the “Google for blockchains”, enables developers like us to tackle these issues.

However, in our testing, we encountered some upcoming limitations as the use of these third party tools and protocols can be slow and requires some manual procedures. Therefore, we decided to build our own Indexer technology, enabling automated token indexing that can be completed within minutes, without any downtime, for thousands of tokens in parallel.

Also, we realized that many Web3 applications were having trouble interacting with blockchain through third-party services,  so we decided to improve this as well:

Blockchain connector (Remote Procedure Call) 

A RPC is a set of protocols and interfaces that enables a software to interact with the blockchain, to read and write, similar to a website interacting with a server. In blockchain terms, a server is referred to as a node, a synced database of the blockchain. Most applications use Infura to do so; it provides RPCs and AWS-like services for nodes, so developers don’t need to build a RPC or manage their own nodes.

Due to Infura’s popularity, blockchain applications become less decentralized and it could become a single point of failure, such as the time it went down because of a service outage one week ago, causing many dApps, including MetaMask, to not function.

It became obvious to us that we had to run our own nodes and build our own RPC to provide better services for our customers. As a result, we have built a system that we can monitor, scale, and rapidly fix if needed. We even enabled switchable RPCs to keep the T-REX Platforms always available by monitoring the state of nodes to connect with functional nodes including our own nodes and third party nodes (e.g. Infura, Alchemy, …).

We are proud of our tech team, especially our DevOps Engineer Cyrille, Lead Core Developer José and Senior Full Stack Developer Mihalis, for their hard work to complete these improvements. We will continue pushing the limit to provide the most advanced, flexible and secure solutions for you.