Construction and Operation of Distributed Price Oracle Machine System

In Oracle Machine services, single points of failure are a common issue. To address this problem, a distributed price Oracle Machine system has emerged. Taking the BTC to USD price service as an example, this system integrates 31 independent price Oracle Machines to collectively provide users with more reliable data services.

The contract design of this aggregation system cleverly addresses the risks that may arise from a single data source. By calling specific methods within the contract, all participating off-chain Oracle Machines can be viewed. Each off-chain Oracle Machine can provide price data through designated methods in response to user requests in the aggregator. These Oracle Machines are typically externally owned accounts (EOA), and they not only provide data for BTC/USD but may also serve other price pairs such as ETH/USD simultaneously.

On-chain contracts follow a strict process when handling data: first, they read the current state and perform a series of checks, and then necessary preparations are made. Next, the contract verifies each signed data to ensure that the signers have the appropriate permissions and prevents double signing. Finally, the contract sorts the observations, selects the median, and ensures that the value is within the preset threshold range.

To further enhance the reliability of the data, the system has also introduced additional verification mechanisms. For example, in certain application scenarios, the obtained prices will be compared with data from other independent sources to ensure that deviations are within an acceptable range.

In addition, to simplify the process of obtaining different token prices, developers have designed the Feed Registry system. This system can be understood as an aggregator of multiple price data sources, allowing users to directly read various token price data through it without needing to separately configure each price data source.

In actual operation, the generation of price data undergoes multiple layers of aggregation. First, there is aggregation at the data source level, collecting raw price data from major trading platforms. Then, there is aggregation at the node operator level, where each operator obtains and processes data from multiple independent data aggregation service providers. Finally, there is aggregation at the entire Oracle Machine network level, usually using the median method to calculate the final price after receiving a sufficient number of node responses.

It is worth noting that not every calculation result will be updated on the chain. The system has set two triggering parameters: deviation threshold and heartbeat threshold, and updates will only occur when one of them is met. Although this mechanism ensures the stability of the data, it also leads to slower price update speeds, which can range from a few minutes to 24 hours. Therefore, this Oracle Machine system is more suitable for application scenarios where real-time price requirements are not high.