Workload Discovery via Real-Time Stream Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large enterprises face challenges in accurately defining application usage and access patterns, leading to difficulties in identifying the smallest set of related data objects for point-in-time consistency in Active/Active continuous availability environments and other software replication scenarios.
Innovation Solution
A method for workload discovery using real-time analysis of input streams, involving the storage of data object changes in a recovery log, analysis by an analytics engine to identify associations based on usage and access patterns, and the subsequent identification of sub-workloads that form consistency groups for replication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a single workload encompassing all data objects is defined, then ease of operation is improved, but manufacturing precision deteriorates
Solution Approach 1:
The patent segments a single large workload into multiple smaller consistency groups based on data object relationships and access patterns. The system automatically divides the monolithic workload into granular consistency groups that can be independently replicated, resolving the contradiction by providing both operational simplicity (automatic segmentation) and precision (granular control over which data objects belong together for consistency).
2Measurement precision
If real-time analysis of replication change streams is performed, then measurement precision is improved, but use of energy deteriorates
Solution Approach 1:
The patent creates a copy of the replication change stream for analysis purposes, allowing the analytics engine to examine usage and access patterns without impacting the primary replication process. This copying approach enables precise measurement of data object relationships while isolating the computational overhead from the critical replication path, thus achieving measurement precision with controlled energy consumption.
3Productivity
If granular consistency groups are identified, then productivity is improved, but device complexity deteriorates
Solution Approach 1:
The system implements self-service by automatically identifying consistency groups and sub-workloads through real-time analysis of replication change streams. The analytics engine autonomously discovers data object relationships and partitions workloads without requiring manual intervention or complex configuration, thereby achieving granular consistency group identification that improves replication productivity while keeping the user-facing complexity low.
Data Source
AI summary
Provided are techniques for workload discovery using real-time analysis of input streams. For a meta workload, changes to data objects made by change operations that are in a replication change stream are stored into a recovery log. Using an analytics engine, one of the recovery log and the replication change stream are analyzed to identify associations between the data objects based on usage and access patterns. The associations are used to identify sub-workloads of the meta workload that form consistency groups for replication.