Cloud Storage Data Replication With Transaction-Aware Failover
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional database management systems face scalability limitations due to bottlenecks in shared-disk architectures and are costly and susceptible to data loss during disasters, while existing data replication strategies like snapshot and logging are impractical for large databases.
Innovation Solution
A hybrid replication strategy combining snapshots with transaction logs to ensure consistent and up-to-date data replication across multiple deployments, enabling failover and failback between primary and secondary deployments, particularly in cloud-based storage systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional shared-disk database architecture is used, then data storage and access is provided, but scalability is limited due to bottlenecks
Solution Approach 1:
The patent segments the database architecture by separating compute nodes from storage nodes. Compute nodes handle query processing while storage nodes manage data persistence, eliminating the shared-disk bottleneck and enabling independent scaling of each component.
Solution Approach 2:
The patent introduces a new architectural dimension by implementing a distributed file system layer between storage nodes and compute nodes. This layer provides abstracted data access and enables scalability without increasing architectural complexity.
2Reliability
If traditional database infrastructure is provisioned in data center, then data storage and management is achieved, but cost is very high
Solution Approach 1:
The patent implements automated data replication across multiple storage nodes and geographic regions. This copying mechanism ensures data safety and disaster recovery while reducing the need for expensive redundant hardware infrastructure.
Solution Approach 2:
The patent creates a multi-functional database system that can operate across different cloud providers and infrastructure types. This universality reduces deployment costs by eliminating vendor lock-in and enabling flexible resource allocation.
3Productivity
If traditional database system is used, then data storage is provided, but system is highly susceptible to data loss during power outage or disaster
Solution Approach 1:
The patent implements preliminary data replication and backup mechanisms that activate before disasters occur. Data is continuously replicated to remote storage nodes and cloud services, ensuring data accessibility during power outages or disasters without impacting normal operations.
Solution Approach 2:
The patent establishes beforehand cushioning through automated failover mechanisms and distributed replication. When disasters occur, pre-configured backup systems and alternative data paths are already in place, preventing data loss and maintaining accessibility.
4Reliability
If snapshot replication strategy is used for data replication, then data backup is achieved, but it is impractical for large databases due to storage overhead
Solution Approach 1:
The patent extracts only the essential data components for replication by implementing incremental replication mechanisms. Instead of copying entire database snapshots, the system replicates only changed data blocks, significantly reducing storage overhead while maintaining data consistency.
Solution Approach 2:
The patent applies partial replication by selectively copying only the portions of data that have changed since the last replication event. This partial action approach maintains data consistency while minimizing storage resource consumption for large databases.
Data Source
AI summary
Replication and failover of data is disclosed. A method includes replicating the data stored in a primary deployment such that the data is further stored in a secondary deployment. The method includes executing one or more transactions on the data at the primary deployment to cause a change to the data to be stored in the primary deployment. The method includes propagating the one or more transactions on the data to the secondary deployment to perform a failover to the secondary deployment while the change to the data is absent from the secondary deployment.