Cluster Database LSN Tracking for Safe Primary Promotion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In cluster management database systems, ensuring database integrity and preventing stale databases from becoming primary instances during instance shutdowns and restarts is challenging, especially when an external witness service is unavailable or fails to provide a quorum.
Innovation Solution
Implementing a system where storage appliances maintain a last sequence number (LSN) for each write operation and store it externally, allowing the most up-to-date appliance to be promoted to primary instance upon restart, while preventing stale databases from being promoted.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a quorum of databases must be up and running to determine the most up-to-date database for promotion, then database integrity is maintained, but system availability deteriorates when appliances are down
Solution Approach 1:
The system performs preliminary actions by writing the LSN to the external witness device before the primary instance shuts down. This advance preparation ensures that when the system restarts with only a subset of appliances, the witness device already contains the necessary information to determine which database is most up-to-date, eliminating the need for a quorum and enabling immediate promotion of the correct instance.
2Productivity
If an external witness service is used to provide an extra vote towards cluster quorum, then quorum is easier to achieve, but the database may still not be promotable without possibility of data loss
Solution Approach 1:
The external witness device serves as an intermediary that stores the LSN information from the primary instance before shutdown. This mediator allows the system to determine database freshness without requiring the primary instance to be present, enabling safe promotion of the most up-to-date secondary instance even when the original primary is down, thus resolving the contradiction between achieving quorum and ensuring data integrity.
3Ease of manufacture
If the primary instance shuts down for updates or maintenance, then the primary can be updated, but the system must wait for quorum before promoting a secondary instance
Solution Approach 1:
By writing the LSN to the external witness device before the primary instance shuts down, the system prepares all necessary information in advance. This allows secondary instances to be promoted immediately upon startup by comparing their LSNs with the witness device, eliminating the waiting period for quorum and reducing overall downtime while still maintaining data integrity.
Data Source
AI summary
A system and method for maintaining database integrity. A first storage appliance receives data writes and increments an associated last sequence number (LSN) X upon each data write thereto. A second storage appliance receives data writes and increments an associated LSN Y upon each data write thereto and an external device is coupled to the first storage appliance and the second storage appliance. The first storage appliance and the second storage appliance comprise a cluster management database with data writes being replicated on both the first storage appliance and the second storage appliance. Upon the first storage appliance beginning to shut down, the LSN X is written to the external device.


