Database Cluster Backup Automation via Primary Instance Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current backup utilities for clustered database systems, such as Microsoft SQL Server AlwaysOn, create redundant backups and make it impossible to achieve an any-point-in-time restore due to multiple versions of databases being maintained under different backup instance contexts.
Innovation Solution
An improved backup utility that automatically detects and backs up the preferred database instance within a clustered database system, storing backups from both primary and secondary nodes under a single backup client, using a constant database instance identifier, allowing for seamless backup and restoration across different nodes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If backup is run on each cluster instance, then each instance is backed up individually, but multiple redundant backups of the same data are created wasting storage space
Solution Approach 1:
The patent merges backup operations across multiple cluster instances by identifying the primary instance and consolidating backups at the cluster level. Instead of creating separate backups for each instance, the system combines them into a single backup operation that captures the entire database group, eliminating redundant storage of identical data across multiple instances.
Solution Approach 2:
The backup system is designed to operate universally across the entire database cluster rather than being instance-specific. The backup utility identifies the primary instance dynamically and performs backups that are valid for the whole cluster, making the backup mechanism multi-functional and applicable to all instances without creating redundant copies.
2Reliability
If backup is run on each cluster instance, then each instance has its own backup, but restoring the group as a whole is impossible
Solution Approach 1:
The patent merges backup operations across multiple cluster instances by identifying the primary instance and consolidating backups at the cluster level. Instead of creating separate backups for each instance, the system combines them into a single backup operation that captures the entire database group, eliminating redundant storage of identical data across multiple instances.
Solution Approach 2:
Instead of backing up each instance separately and trying to reconstruct the group, the patent inverts the approach by backing up the group as a single unit through the primary instance. This inversion enables group-level restoration while maintaining instance-level backup integrity, as the primary instance contains the authoritative copy of the database group state.
3Productivity
If backup utility backs up from different nodes, then backups are taken from current primary node, but multiple versions are maintained under different backup instance contexts
Solution Approach 1:
The backup system is designed to operate universally across the entire database cluster rather than being instance-specific. The backup utility identifies the primary instance dynamically and performs backups that are valid for the whole cluster, making the backup mechanism multi-functional and applicable to all instances without creating redundant copies.
Solution Approach 2:
The backup utility implements feedback mechanisms to track the primary instance identity and backup state across cluster operations. By monitoring failover events and primary instance changes, the system automatically adjusts backup operations to maintain a single version of the database group, preventing the creation of multiple versions under different contexts.
4Adaptability or versatility
If manual backup preference setting is required, then user can control backup node, but backup process becomes complex and error-prone
Solution Approach 1:
The backup system performs self-service by automatically identifying the primary instance and configuring backup operations without requiring manual user input. The backup utility monitors the cluster state, detects the current primary instance, and executes backups autonomously, eliminating the need for manual preference setting while maintaining flexibility through automatic adaptation to cluster changes.
Data Source
AI summary
An improved backup utility allows the ability to maintain full backup and incremental backup log chains, even when databases which are part of a clustered database group are backed up from different database nodes and instances. In these embodiments, this is implemented by storing the backup of the databases from both primary and secondary (failover) nodes under a single backup client which is the cluster identifier. The backup is indexed against a constant database instance based on a user configured save set. In other words, regardless of which node is primary at the time of the backup, the backup is made against a constant backup instance.