System risk control mechanism

CoinConnectPro puts security first, so our system risk control mechanism startsfrom the following aspects:

1.Database read-write separation mechanism

In the initial stage,the system risk control generallyensures thesynchronization and read-write separation between the database of the trading system and the risk control system through the establishment of database master-slave replication, read-write separation, Sharding and other mechanisms.System risk control generally only has the permission to read the required customer/account data and transaction data, so as to ensure the safety and reliability of account data.

2.Cache/memory database mechanism

An efficient cache system is an effective measure to improve performance, andgenerally the mechanism will store frequently used data in a cache system such asRedis. For example, risk control rules, risk control case base, intermediate result set.black and white list, pre-processing results, transaction parameters, billing templateclearing and settlement rules, distribution rules and other data. For somehigh-frequency transactions, based on performance:considerations,memorydatabases will be used for storage (generally combined with ssD disks).

3.RPC/SOA architecture

Reduce the coupling of trading system and system risk control. in the initialcase of less system services, the general direct use of RabbitMQ/ActiveMQ andother message-oriented middleware or RPC mode to achieve the inter-systemservice call. When the number of system services increases and service governanceproblems occur, SOA middleware such as Dubbo will be used to realize the systemservice invocation.

4.Composite Event Processing (CEP)

Real-time/quasi-real-time trading risk control. Compared with the purerule-based processing mode, the composite event processing (CEP) mode hasbetter performance and scalability.