Standby Database Read Availability via Mediator Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Physical standby database systems are unavailable for read operations until the application of redo records is stopped, leading to delays and potential data corruption or loss, and similar limitations exist in remote mirroring systems where read operations on replicas are delayed due to updating delays.
Innovation Solution
Implement mechanisms that allow changes from a primary database system to be applied to replicas while enabling concurrent read operations by maintaining a latest-readable query time and ensuring cache coherency, including consistent read approaches and lazy invalidation of meta-data caches, allowing for simultaneous data replication and read access.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If read operations are enabled on physical standby database system, then availability for reporting applications is improved, but data consistency and reliability deteriorate due to potential corruption from concurrent updates
Solution Approach 1:
The patent introduces a mediator mechanism that intercepts and manages write operations to the standby database. The mediator captures update operations, applies them to a copy of the data, and maintains a mapping between primary and standby data blocks. This intermediary layer allows read operations on the standby while protecting data consistency by controlling how updates are propagated.
Solution Approach 2:
The patent segments the database system into distinct components: primary database system, standby database system, and a coordination mechanism. The data is divided into data blocks with corresponding mappings between primary and standby systems. This segmentation allows independent management of read and write operations on different segments, enabling reads on standby while updates are coordinated through the mapping mechanism.
2Productivity
If read operations are performed on physical standby database system, then system utilization is improved, but update delays increase leading to data lag
Solution Approach 1:
The patent implements preliminary action by pre-establishing a mapping between primary and standby data blocks before read operations begin. The mapping is created in advance, allowing the system to quickly determine which standby blocks correspond to primary blocks without delay. This preliminary setup enables concurrent reads and updates without time loss.
Solution Approach 2:
The patent introduces dynamic adjustment of the mapping mechanism. The mapping between primary and standby data blocks is not static but can be dynamically updated and adjusted as data changes. This dynamic approach allows the system to adapt to changing data states while maintaining efficient read access, preventing update delays even as data evolves.
3Ease of operation
If redo record application is stopped to enable reads on standby, then read availability is improved, but data freshness deteriorates
Solution Approach 1:
The patent enables continuity of useful action by allowing redo record application to continue concurrently with read operations on the standby database. The mapping mechanism ensures that updates from redo records are properly tracked and propagated while reads can simultaneously access the standby data. This continuous operation maintains both data freshness and read availability without interruption.
Data Source
AI summary
A standby database system or another replica data system replicates changes, made to data blocks at a source database system or another primary data copy, to replicas of the data blocks at the standby database system or other replica. While replicating the changes to the data blocks thereof, the standby database system (or other replica) receives queries (or reads) issued thereto and computes the queries based on data read from the data blocks thereof.