Object Tiering System Bundles Data for Cloud Transfer
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current virtual computing systems face inefficiencies in resource utilization and performance due to their configuration and operation, particularly in managing data storage across local and cloud environments.
Innovation Solution
An object tiering system that ingests data objects, maps them into regions based on commonality such as tiering dates, and transfers these regions efficiently from a local store to a cloud store, enabling flexible and scalable storage management by sharding data and metadata across multiple clouds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Volume of stationary object
If data objects are transferred individually from local store to cloud store, then storage capacity is increased, but transfer time and system complexity increase
Solution Approach 1:
The patent combines multiple data objects into a single data bundle for transfer. The system identifies multiple data objects eligible for cloud storage, groups them together into one bundle, and transfers the entire bundle in a single operation. This merging approach maintains the benefit of increased storage capacity while dramatically reducing transfer time by eliminating repeated individual transfer operations.
Solution Approach 2:
The system performs preliminary identification and grouping of data objects into bundles before the actual transfer to cloud storage occurs. By pre-processing the data objects and organizing them into bundles in advance, the system prepares the transfer workload optimally, reducing the time required during the actual transfer phase and improving overall efficiency.
2Volume of stationary object
If data objects are transferred individually from local store to cloud store, then storage capacity is increased, but device complexity increases
Solution Approach 1:
The patent combines multiple data objects into a single data bundle for transfer. The system identifies multiple data objects eligible for cloud storage, groups them together into one bundle, and transfers the entire bundle in a single operation. This merging approach maintains the benefit of increased storage capacity while dramatically reducing transfer time by eliminating repeated individual transfer operations.
Solution Approach 2:
The system implements a universal data bundle structure that can accommodate multiple different data objects with varying characteristics. This multi-functional bundle approach allows the same transfer mechanism to handle diverse data types and sizes uniformly, simplifying the system architecture by providing a single standardized interface for data ejection rather than requiring separate handling logic for each individual object.
3Adaptability or versatility
If data is sharded across multiple clouds, then storage flexibility is improved, but management complexity increases
Solution Approach 1:
The patent implements segmentation by dividing data into bundles that can be distributed across multiple cloud storage locations. Each data bundle can be independently managed and placed in different cloud environments based on requirements. This segmentation provides the flexibility to shard data across multiple clouds while maintaining manageable units that can be handled independently, reducing the complexity of managing distributed data across multiple cloud providers.
Data Source
AI summary
An object tiering system transfers or tiers data objects from a data store to a cloud store. The system includes a processor and a memory having stored instructions that, when executed by the processor, cause the processor to ingest a first plurality of data objects in the data store, map a second plurality of data objects having a common tiering date to a region, and upon reaching the common tiering date, transfer the region to the cloud store. In some embodiments, the mapping is done a priori when each data object is ingested or at a first time before the common tiering date. In some embodiments, the common tiering date is a second time for transferring the second plurality of data objects from the data store to the cloud store. In some embodiments, the first plurality of objects includes the second plurality of objects.