RE-aware Slot Placement for Data Center Bandwidth Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing virtual machine placement strategies in data centers do not effectively minimize bandwidth requirements across racks, leading to inefficient data transfer and bandwidth usage.
Innovation Solution
An RE-aware technique for placing slots based on redundancy across and within slot communication pairs, which involves determining redundancy parameters and optimizing slot placement within containers to reduce inter-container bandwidth usage by exploiting intra-comm-redundancy and inter-comm-redundancy patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If virtual machines are placed in data center racks based on minimizing bandwidth requirements of data transfers across racks, then bandwidth usage is reduced, but placement optimization is insufficient due to lack of redundancy awareness
Solution Approach 1:
The patent changes the placement parameters from traditional bandwidth-only considerations to include redundancy parameters (intra-comm-redundancy and inter-comm-redundancy). This allows the system to optimize placement by considering both bandwidth requirements and redundancy patterns, achieving better bandwidth utilization while maintaining manageable optimization complexity through structured parameter expansion.
Solution Approach 2:
The patent segments the redundancy analysis into two distinct components: intra-comm-redundancy (within slot communication pairs) and inter-comm-redundancy (across slot communication pairs). This segmentation enables the system to handle complex redundancy patterns systematically, improving placement optimization without overwhelming complexity by breaking down the overall optimization problem into manageable parts.
2Ease of manufacture
If traditional bandwidth-based placement strategies are used, then implementation is simple, but bandwidth usage is not optimized due to ignoring redundancy patterns
Solution Approach 1:
The patent performs preliminary redundancy measurement and parameter determination before slot placement occurs. By measuring redundancy patterns (intra-comm and inter-comm) in advance and using these pre-computed parameters to guide placement decisions, the system achieves optimized bandwidth usage while maintaining relatively simple implementation. The preliminary action of redundancy measurement enables subsequent placement optimization without adding complex real-time computation during placement operations.
3Speed
If slots are placed without considering redundancy patterns, then placement process is fast, but inter-container bandwidth usage is not minimized
Solution Approach 1:
The patent computes redundancy parameters (intra-comm-redundancy and inter-comm-redundancy) in advance before the actual slot placement process. This preliminary computation of redundancy patterns allows the placement algorithm to quickly use pre-computed information rather than performing complex redundancy analysis during placement, thereby maintaining fast placement speed while achieving optimized inter-container bandwidth usage through the pre-determined redundancy parameters.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Various embodiments provide a method and apparatus of providing an RE-aware technique for placing slots based on redundancy across and within slot communication pairs.