Object Store Point-in-Time Recovery Architecture
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current storage systems face challenges in efficiently replicating and recovering data to a remote or cloud-based replication site, particularly in identifying and restoring data to a specific point in time for disaster recovery or data access.
Innovation Solution
A storage system architecture that includes a production site and a cloud replication site, where an object store generates and manages multiple points in time (PITs) by storing data objects and their changes, allowing for the identification and recovery of specific PITs through virtual machines that process disk, change, and metadata objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is replicated to a remote or cloud-based replication site, then data protection and availability are improved, but storage requirements and compute power usage increase
Solution Approach 1:
The patent segments data into discrete objects with associated metadata, allowing selective replication and storage of only necessary data portions. Each data object is independently managed, enabling efficient use of storage resources while maintaining data protection across multiple replicas.
Solution Approach 2:
The object store architecture provides multi-functional capabilities including data replication, versioning, recovery, and access management within a single system. This universal approach consolidates multiple data protection functions, reducing overall storage requirements compared to separate systems for each function.
2Adaptability or versatility
If multiple points in time are generated and stored, then data recovery flexibility is improved, but storage requirements increase
Solution Approach 1:
The patent implements a nested structure where data objects contain embedded metadata including version information and point-in-time identifiers. This nesting allows multiple temporal versions to be stored efficiently within the same object hierarchy, providing recovery flexibility without proportionally increasing storage requirements.
Solution Approach 2:
The system uses parameter changes in metadata (timestamps, version numbers, PIT identifiers) to differentiate data states over time without duplicating entire data sets. By changing metadata parameters rather than storing complete copies, the system achieves flexible point-in-time recovery with optimized storage usage.
3Reliability
If data objects are replicated across multiple sites, then data availability is improved, but compute power usage increases
Solution Approach 1:
The patent performs preliminary actions by pre-generating and storing multiple points in time (PITs) and data replicas in advance before recovery is needed. This preliminary replication and organization of data objects reduces the compute power required during actual recovery operations, as data is already prepared and positioned for rapid retrieval.
Data Source
AI summary
Described embodiments provide devices, systems and methods for operating a storage system. An object store located at the replication site stores data objects associated with data stored in storage of the production site. The replication site may generate a plurality of points in time (PITs) from the data objects and identify a PIT from the plurality of PITs.