Cloud Resource Allocation via Dynamic Time Warping Stretch Factors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cloud computing environments face challenges in efficiently allocating resources across multiple hosts due to varying hardware configurations and the difficulty in identifying running services without intrusive methods.
Innovation Solution
A method that identifies services running on multiple hosts, determines a stretch factor for recurring load patterns using Dynamic Time Warping (DTW), and stores the service with its stretch factor, enabling non-intrusive resource allocation and classification across different hardware configurations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If virtualization is used to segment resources, then resource usage efficiency is improved, but virtualization overhead reduces performance
Solution Approach 1:
The system dynamically selects between virtualization and containerization based on workload characteristics. For workloads requiring high performance, containerization is used to minimize overhead. For workloads requiring strong isolation, virtualization is applied. This dynamic approach optimizes the balance between resource efficiency and performance.
Solution Approach 2:
The patent changes the isolation parameter from always using virtualization to conditionally using either virtualization or containerization based on service requirements. This parameter change allows the system to adjust the level of isolation and overhead according to specific workload demands.
2Adaptability or versatility
If hardware configurations vary across hosts, then system adaptability is improved, but service identification accuracy deteriorates
Solution Approach 1:
The patent introduces an intermediary layer that abstracts service identification from hardware-specific characteristics. Instead of directly monitoring hardware parameters, the system uses intermediary metrics such as process behavior patterns, resource usage profiles, and service-level indicators that remain consistent across different hardware configurations.
Solution Approach 2:
The system creates a standardized representation or copy of service characteristics that is independent of the underlying hardware. By copying and normalizing service profiles across different hosts, the system maintains consistent identification accuracy regardless of hardware variations.
3Ease of operation
If non-intrusive monitoring is used for service identification, then system interference is reduced, but monitoring precision deteriorates
Solution Approach 1:
The system implements feedback mechanisms where initial non-intrusive monitoring results are continuously refined based on service behavior patterns and historical data. The monitoring system learns from feedback loops, improving precision over time while maintaining non-intrusive operation.
Solution Approach 2:
The patent performs preliminary service identification using non-intrusive methods, then uses the initial results to guide more precise monitoring strategies. By performing preliminary classification, the system can apply targeted monitoring only where needed, maintaining precision without excessive interference.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach allows for accurate identification of services across different hardware, enabling optimized resource allocation and cost management in cloud environments by predicting load patterns and matching resources with workload demands.
Implementation Method 1
determining a stretch factor for a recurring load pattern for the service running on the one or more of the plurality of hosts
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
It is presented a method for enabling allocation of resources for a plurality of hosts. The method is performed by a server (l) and comprises identifying (S100) a service running on one or more of the plurality of hosts, determining (S140) a stretch factor for a recurring load pattern for the service running on the one or more of the plurality of hosts, and storing (S150) the identified service together with the determined stretch factor. It is also presented a server, a computer program and a computer program product.