Lease-Based Consistency Management for Database Failover
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Distributed database systems face disruptions due to network and other failures, leading to inconsistent data access and potential stale data issues, particularly when failovers occur without proper consistency management.
Innovation Solution
Implementing lease-based consistency management, where database nodes independently determine their roles using consistency leases to ensure consistent access, allowing each node to detect and adapt to failovers, thereby preventing inconsistent reads and writes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If distributed database systems are implemented to scale processing capacity and increase service availability, then productivity and reliability are improved, but data consistency and access reliability deteriorate due to network failures and failover scenarios
Solution Approach 1:
The system performs preliminary actions by establishing consistency leases before failover occurs. The primary database node obtains a consistency lease that spans the failover period, and read-only nodes proactively detect failover and obtain new leases before actual failover happens, preventing inconsistent data access
Solution Approach 2:
The system implements feedback mechanisms where read-only database nodes continuously monitor for failover events and attempt to obtain consistency leases. When a node successfully obtains a lease, it provides feedback to other nodes, enabling them to update their state and maintain consistent data access across the distributed system
2Reliability
If failover mechanisms are implemented to handle network failures, then reliability is improved, but data consistency deteriorates leading to potential stale data issues
Solution Approach 1:
The system establishes consistency leases that cover the failover period in advance. The primary node's lease spans the expected failover window, and read-only nodes proactively attempt to obtain leases before failover occurs, ensuring they have valid leases and consistent data views when failover actually happens
Solution Approach 2:
The consistency lease acts as an intermediary mechanism that mediates between the primary and read-only database nodes during failover. The lease provides a shared reference that ensures all nodes operate with consistent data views, preventing stale data issues while enabling reliable failover
3Device complexity
If database nodes independently determine their roles without consistency management, then device complexity is reduced, but data consistency deteriorates leading to inconsistent reads and writes
Solution Approach 1:
The consistency lease serves as an intermediary that coordinates role determination across database nodes. Instead of complex inter-node communication protocols, nodes use the lease as a shared reference to independently determine their roles while maintaining consistency, simplifying the system while ensuring reliability
Solution Approach 2:
Database nodes independently obtain and use consistency leases to determine their own roles without requiring complex coordination with other nodes. Each node services itself by monitoring lease status and adjusting its role accordingly, reducing overall system complexity while maintaining access consistency
Data Source
AI summary
Lease-based consistency may be implemented for databases to handle failovers. A database node may obtain a consistency lease that describes a point in time determined from a time-to-live amount added to a consistent point in time for database data. While the consistency lease is valid, Multi-version Concurrency Control (MVCC) snapshots assigned by the database node can be used to handle requests to access the database data. Once expired, the database node may have to renew the consistency lease in order to continue to handle write and read requests.