Description

GitHub: https://github.com/chainstacklabs/unstoppable-rpc

Contact person: Vasily Rudomanov

Discord: cyberwider#7763

Description:

Usual RPC endpoints from Chainstack, Alchemy and Infura are stuck to their infra that leads to lower uptime, higher latency and no custom logic for an RPC endpoint.

What if there will be a service that can aggregate multiple RPC endpoints from different RPC providers, geographical regions, cloud providers and:

1. check if a node is up and available

2. check if a node has the latest block height and there is no headlag

3. check what latency is from an end-user to an RPC node

4. check how requests per second (RPS) that node can serve (how much it is utilized already)

and based on above criteria, provider intelligent re-routing to the best node.

Moreover, for that chosen node, our service will apply:

1. an access-list capabilities to white-list IP-addresses or HTTP Referrer origin value

2. introduce a custom rate-limiting and throttling parameters

3. detect if an RPC request is going for full or archive node, and based on that redirect request to a node that has either only full data (latest blocks) or archival data (all blocks from the genesis block)


Such service will greatly improve:

1.  availability (uptime) of an RPC node

2. with the lowest latency from an end user to a chosen RPC node

3. make sure the node has the latest chain data

4. make sure the node is not over congested and has enough resources to serve your RPC traffic

5. implement the security model so your RPC endpoint will not be accessed by 3rd parties

6. reduce engineering\devops overhead by implementing logic by accessing different RPC endpoints based on block height (a universal endpoint both for latest and archival data)


And this is The Unstoppable RPC Endpoint, ready-to-go service, open sourced and available for everyone.

It helps for every DApp project, for any EVM-comptable chain, and provide unbound user experience for the next billion of web3 users.


Attachments