Cross-Data-Center Resource Allocation for Workload Bottlenecks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In data centers where multiple workloads are executed concurrently, resource allocation challenges arise when a workload requires specific resources that are already allocated to another workload, leading to inefficiencies and performance issues.
Innovation Solution
A data center architecture that disaggregates resources into sleds optimized for specific functions (compute, memory, storage, accelerators) and allocates these resources dynamically to form managed nodes, allowing flexible and efficient utilization across workloads.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If resources are allocated to multiple workloads concurrently in a data center, then resource utilization efficiency is improved, but resource availability for individual workloads deteriorates when resources are already allocated
Solution Approach 1:
The patent segments resources into two categories: local resources within a data center and remote resources in off-premises data centers. This segmentation allows workloads to first utilize local resources and then access remote resources when local resources are exhausted, resolving the contradiction between maximizing local resource utilization and ensuring resource availability for individual workloads.
Solution Approach 2:
The patent creates a universal resource pool that combines local and remote data center resources, allowing any workload to access resources from either location based on availability. This multi-functionality enables the system to maintain high resource utilization efficiency while ensuring resource availability for all workloads simultaneously.
2Reliability
If resources are allocated to satisfy service level agreements, then workload performance is improved, but resource allocation flexibility deteriorates when resources are unavailable
Solution Approach 1:
The patent establishes resource allocation agreements in advance between data centers, creating pre-configured resource pools that can be rapidly deployed when needed. This preliminary action ensures that resources are available to meet service level agreements without requiring complex real-time allocation negotiations, thus maintaining both reliability and flexibility.
Solution Approach 2:
The patent introduces an intermediary orchestration layer that manages resource allocation between local and remote data centers. This intermediary automatically selects and allocates resources from the appropriate data center based on availability and service level requirements, maintaining allocation flexibility while ensuring reliable service delivery.
3Ease of operation
If resources are statically allocated within a single data center, then resource management simplicity is improved, but resource utilization efficiency deteriorates when workloads have varying resource needs
Solution Approach 1:
The patent implements dynamic resource allocation where the system automatically adjusts resource allocation between local and remote data centers based on real-time workload demands and resource availability. This dynamic approach maintains operational simplicity through automated decision-making while significantly improving resource utilization efficiency by allocating resources where they are most needed.
Data Source
AI summary
Technologies for allocating resources across data centers include a compute device to obtain resource utilization data indicative of a utilization of resources for a managed node to execute a workload. The compute device is also to determine whether a set of resources presently available to the managed node in a data center in which the compute device is located satisfies the resource utilization data. Additionally, the compute device is to allocate, in response to a determination that the set of resources presently available to the managed node does not satisfy the resource utilization data, a supplemental set of resources to the managed node. The supplemental set of resources are located in an off-premises data center that is different from the data center in which the compute device is located. Other embodiments are also described and claimed.