Incremental Snapshot Hibernation for Bare Metal Clusters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for hibernating and resuming bare metal clusters in cloud computing environments involve full copy and restore processes, leading to high time and cost inefficiencies due to the size of the data-set, and result in ongoing resource usage costs even when clusters are idle.
Innovation Solution
Implementing incremental hibernate and resume techniques that generate full and incremental snapshots of cluster resources and configuration information, allowing processes to be restored quickly and efficiently, with instant restore capabilities and metadata-driven resource access.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If full copy and restore processes are used for hibernating and resuming bare metal clusters, then complete data preservation is achieved, but hibernate and resume times become very high due to the size of the data-set
Solution Approach 1:
The patent segments the backup process into full snapshots and incremental snapshots. Full snapshots capture the complete cluster state initially, while subsequent incremental snapshots only capture changes since the last snapshot. This segmentation allows the system to preserve complete data (through the combination of full and incremental snapshots) while reducing the time required for hibernation and resumption, as only incremental changes need to be processed after the initial full copy.
Solution Approach 2:
The patent performs preliminary actions by capturing full snapshots before hibernation and maintaining incremental snapshots in advance. When hibernation is needed, the system already has pre-captured snapshot data available, eliminating the need to perform full data copying during the hibernation process itself. This preliminary capture of data states significantly reduces the actual hibernation and resume time while ensuring complete data preservation.
2Reliability
If full copy and restore processes are used for hibernating and resuming bare metal clusters, then complete data preservation is achieved, but resource usage costs increase due to ongoing charges during hibernation and resume periods
Solution Approach 1:
By segmenting the backup into full and incremental snapshots, the patent reduces the amount of data that needs to be processed during hibernation and resume operations. This allows the cluster to be hibernated more quickly, reducing the time resources remain active and incurring costs, while still ensuring complete data preservation through the stored snapshot sequence.
Solution Approach 2:
The patent uses snapshot copying to create point-in-time representations of the cluster state without requiring full data duplication during hibernation. The full snapshot captures the complete state initially, and incremental snapshots copy only the changes. This copying mechanism enables rapid hibernation and resume by working with compact snapshot data rather than full dataset copies, reducing resource usage time and cost while maintaining data integrity.
3Loss of time
If incremental snapshot generation is used for quick hibernation, then hibernate and resume times are reduced, but system complexity increases due to snapshot management requirements
Solution Approach 1:
The patent implements self-service mechanisms where the snapshot management system automatically handles the complexity of full and incremental snapshot generation, storage, and restoration. The system autonomously tracks which snapshots exist, which are full versus incremental, and coordinates their application during resume operations. This automation reduces the operational burden on users despite the underlying complexity of managing incremental snapshot sequences.
Solution Approach 2:
The patent introduces an intermediary snapshot management layer that sits between the cluster data and the hibernation/resume operations. This intermediary component handles the complexity of coordinating full and incremental snapshots, managing their storage and retrieval, and orchestrating the restore process. By placing this intermediary in charge, the system manages snapshot complexity centrally rather than requiring distributed complexity throughout the cluster, reducing overall system complexity while enabling fast incremental hibernation.
Data Source
AI summary
The present disclosure describes techniques used in systems, methods, and in computer program products to perform incremental hibernate and resume of bare metal clusters. A replication process can generate full/incremental snapshots at the cluster level and at a resource level of the processes of the cluster. In the even that a user instructs the cluster to hibernate some or all processes of the cluster, the process can capture an incremental snapshot of the corresponding resources far more quickly than would be required to generate full snapshots for those resources. Likewise, system configuration information can be captured to enable the underlying infrastructure of the cluster to be restored in the future. In some embodiments, the approach provided herein also implements an instant restore feature.