Taking a closer look at the architecture of Chainlink oracles, it’s important to understand that there are two different models.
The current deployment of Chainlink oracles on the Stacks testnet uses the ‘Direct Request Model
’. Essentially, developers can get real-world data into their contract by submitting an on-chain request to a Chainlink node in one transaction and receive the response to the query within a separate transaction. The Direct Request Model is the first oracle model planned to be deployed on the Stacks mainnet to bring real-world data on-chain. After this model has been sufficiently time-tested and level of demand sufficient, the next step is to explore providing support for the ‘Off-Chain Reporting
’ model, otherwise known as OCR.
The Chainlink OCR model enables the creation of on-chain data feeds that are continuously updated by a decentralized network of oracles that collectively fetch and aggregate data off-chain from multiple sources and deliver it on-chain in the form of aggregated oracle reports. Chainlink oracle nodes in the OCR model communicate peer-to-peer directly off-chain, generating consensus on external data, which is delivered to an on-chain reference contract in a single transaction for each round. This model is ideal for obtaining freshly updated real world data within a single on-chain query, so as a community we want to work toward this implementation by proving demand for the Direct Request Model now.