Database Replication Using Cyclic Redundancy Check Tree Indexes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing database replication methods in process information management systems face challenges in maintaining high availability and performance, especially in mission-critical environments like oil refineries and nuclear power plants, as they are prone to single-point failures and do not efficiently handle relational data redundancy.
Innovation Solution
A method and system for replicating databases using a B-tree or B+ tree index structure with cyclic redundancy check values, which ensures logical consistency by comparing sums of these values across databases, dividing tree indexes as needed to identify and replicate inconsistent data, allowing for uninterrupted operation and high availability without the need for transaction archives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If PC clustering technology is used to build redundant process information management systems, then system availability is improved through failover capability, but system operation experiences undesirable breaks during failure transitions
Solution Approach 1:
The system performs preliminary actions by continuously maintaining synchronized database copies across cluster nodes before failures occur. The standby node is pre-configured with up-to-date database state, enabling immediate takeover without operational breaks when a primary node fails.
Solution Approach 2:
The invention creates and maintains copied database instances across multiple cluster nodes. Each node holds a copy of the database, allowing any node to immediately assume the primary role without data reconstruction or system interruption when a failure occurs.
2Reliability
If traditional database replication methods are used in process information management systems, then data redundancy is achieved, but performance requirements for real-time operation are not met
Solution Approach 1:
The invention segments the database into fixed-size pages and uses a B-tree index structure divided into manageable blocks. This segmentation enables efficient parallel processing and reduces the overhead of replication operations, meeting real-time performance requirements while maintaining data redundancy.
Solution Approach 2:
The system changes parameters by using fixed 4KB page sizes and structured B-tree blocks with predetermined organization. This parameter standardization optimizes storage efficiency and access performance, enabling real-time operation while maintaining redundant copies across nodes.
3Reliability
If process historian products are used to provide redundancy for process history, then historical data redundancy is achieved, but capability to store and manage relational data is lost
Solution Approach 1:
The invention creates a universal database system that can simultaneously store and manage both process history data and relational data with full SQL capability. The same clustered database architecture handles multiple data types and query requirements, eliminating the need for separate process historian products.
Solution Approach 2:
The B-tree index structure with cyclic redundancy check sums acts as an intermediary mechanism that enables efficient data verification and replication across cluster nodes. This intermediary structure supports both historical data redundancy and relational data integrity without requiring separate systems.
4Measurement precision
If B-tree indexes with cyclic redundancy check sums are used for database replication, then data consistency is efficiently verified across distributed databases, but index structure complexity increases
Solution Approach 1:
The invention replaces complex mechanical data comparison mechanisms with a mathematical approach using cyclic redundancy check sums. Instead of comparing individual data records, the system computes and compares cryptographic hash values of entire B-tree blocks, dramatically simplifying consistency verification across distributed databases.
Solution Approach 2:
The B-tree index structure with embedded cyclic redundancy check sums enables self-verification of data consistency. Each node can independently verify its data副本 against other nodes by comparing hash values without requiring complex external verification protocols, reducing overall system complexity.
Data Source
Figure 1~1C
Figure 2
Figure 3
AI summary
A method for replicating databases, in which method at least two databases are replicated, each database comprising of one or more tables, and each table having data and a unique tree index comprising index keys and hierarchical sums of cyclic redundancy check values calculated from the data on each tree index level. In the method logical consistency of the data between corresponding tables in the databases is maintained by comparing the sums of cyclic redundancy check values of the database tables; and if a difference between the sums of the cyclic redundancy check values is found, the tree indexes are logically divided into two sub tree indexes, and the sums of the cyclic redundancy check values of the sub tree indexes are compared to each other and the comparison and division is continued until the data causing the difference is found, and the inconsistent data is replicated between the databases.