Workload Telemetry Fingerprinting for Dynamic Resource Allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In computing environments, provisioning maximum resources for workloads can lead to overprovisioning and waste due to varying resource requirements across different segments of a workload, such as memory-intensive or CPU-intensive segments.
Innovation Solution
A method using fingerprint representations of telemetry data to identify workload segments and apply appropriate resource allocation policies, leveraging a fingerprint catalog to optimize resource utilization by matching workload segments with similar resource requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If maximum resources are provisioned for the entire workload execution time, then the workload can be executed without resource constraints, but compute resources are overprovisioned during periods when less resources are needed
Solution Approach 1:
The workload execution timeline is segmented into distinct phases based on telemetry data analysis. Resource allocation policies are applied separately to each phase, allowing resources to be adjusted according to the specific needs of each segment rather than maintaining maximum resources throughout the entire execution period.
Solution Approach 2:
Resource allocation transitions from a static maximum provisioning approach to a dynamic approach where resource levels are adjusted based on the current workload phase. The system continuously monitors telemetry data and modifies resource allocation in real-time to match actual workload requirements, preventing both overprovisioning and underprovisioning.
2Productivity
If resource allocation is adjusted dynamically based on workload segments, then resource utilization efficiency improves, but system complexity increases due to multiple policies and telemetry monitoring
Solution Approach 1:
Resource allocation policies for different workload phases are predetermined and stored in a policy database before workload execution begins. When telemetry data indicates a specific phase is active, the corresponding pre-defined policy is automatically selected and applied, eliminating the need for complex real-time decision-making algorithms while maintaining dynamic adaptability.
Solution Approach 2:
The system implements continuous feedback loops where telemetry data from workload execution is monitored and fed back to the resource allocation mechanism. This feedback drives automatic policy selection and resource adjustment, creating a closed-loop control system that adapts to changing workload conditions without requiring manual intervention or complex heuristic algorithms.
Data Source
AI summary
Techniques described herein relate to a method for resource allocation using fingerprint representations of telemetry data. The method may include receiving, at a resource allocation device, a request to execute a workload; obtaining, by the resource allocation device, telemetry data associated with the workload; identifying, by the resource allocation device, a breakpoint based on the telemetry data; identifying, by the resource allocation device, a workload segment using the breakpoint; generating, by the resource allocation device, a fingerprint representation using the workload segment; performing, by the resource allocation device, a search in a fingerprint catalog using the fingerprint representation to identify a similar fingerprint; obtaining, by the resource allocation device, a resource allocation policy associated with the similar fingerprint; and performing, by the resource allocation device, a resource policy application action based on the resource allocation policy.