Block Storage Replication Leases for Zone Failure Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current block storage services face issues with zone unavailability due to power outages or failures, leading to data inaccessibility and potential disk corruption from multiple writers, especially in systems with high availability requirements.
Innovation Solution
Implementing synchronous replication of data across two user-specified zones within a single region, using a replicated block storage system that maintains replicas in different zones and employs health state determination, degraded file markers, and replication leases to ensure data durability and availability during zone failures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is replicated synchronously across two zones, then data durability and availability during zone failures is improved, but system complexity and operational overhead increase
Solution Approach 1:
The system segments the storage architecture into distinct zones with independent replicas. Each zone maintains a complete replica of the disk data, allowing the system to partition fault isolation to the zone level. This segmentation enables continued operation from healthy zones while preventing failure propagation across the entire system.
Solution Approach 2:
The patent introduces intermediary components including a coordination service that manages replica health state determination, write command routing, and failover coordination. These intermediaries abstract the complexity of multi-zone replication from individual nodes, centralizing control logic and simplifying operational management despite the distributed architecture.
2Reliability
If synchronous replication is implemented across zones, then protection against disk corruption from multiple writers is improved, but write operation latency increases
Solution Approach 1:
The system performs preliminary health state determination and replica validation before executing write operations. By pre-establishing which replicas are healthy and capable of accepting writes, the system avoids unnecessary write attempts to failed replicas, reducing latency while maintaining data integrity through selective write routing to valid targets.
Solution Approach 2:
The system dynamically adjusts write operation behavior based on real-time replica health states. When replicas become unavailable, the system adapts by routing writes only to healthy replicas and updating coordination information. This dynamic adaptation allows the system to maintain data integrity while optimizing write latency according to current system conditions rather than following rigid synchronous protocols.
3Productivity
If health state determination and degraded file markers are used, then data reconciliation efficiency is improved, but system operational complexity increases
Solution Approach 1:
The system uses degraded file markers as visual indicators (analogous to color changes) to identify files that require reconciliation between replicas. These markers provide an immediate, easily detectable signal of data state discrepancies, allowing operational systems to quickly identify and prioritize reconciliation tasks without complex analysis, thereby improving efficiency while keeping the indication mechanism simple.
Solution Approach 2:
The health state determination mechanism provides continuous feedback to the coordination service and operational systems about replica status. This feedback loop enables automatic detection of degraded states, triggering reconciliation processes only when necessary. The feedback-driven approach improves reconciliation efficiency by eliminating unnecessary operations on healthy replicas while maintaining simple operational procedures through automated status monitoring.
Data Source
AI summary
A replicated block storage service provides durable and high performance network-attached storage replicated in two or more zones of a single region, and remains available despite a single zone failure. A probe file is generated to determine a health state of a replicated disk. When a disk is degraded, a lease is created indicating which replica is trusted and providing visibility to backend jobs to facilitate reconciliation of data between the first replica and the second replica. Moreover, degraded file markers are generated for use by the backend jobs in quickly identifying the data to be copied.