Dynamic Memory Pool Management for Local-Remote Data Placement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cloud computing environments face challenges with high latency in data transfer for time-sensitive applications due to heterogeneous memory pools, leading to increased costs from adding infrastructure to meet demand fluctuations.
Innovation Solution
Implementing a control group API to manage memory allocation across local and remote memory pools, allowing cloud service providers to optimize memory usage based on service level agreements (SLAs) and workload criticality, using AI-based prediction for data placement strategies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If memory capacity is expanded by adding new pools, then storage capacity increases, but total cost of ownership increases and resource utilization efficiency decreases during low demand periods
Solution Approach 1:
The patent implements dynamic memory allocation where memory pages are migrated between local memory and remote memory pools based on real-time access patterns. The system continuously monitors workload demands and adjusts memory placement accordingly, allowing the same physical infrastructure to adapt to varying demand levels without requiring permanent expansion for peak capacity
Solution Approach 2:
The patent creates a universal memory management system that serves multiple functions: it provides fast local memory access for frequently used data, offers expanded remote storage for less frequently accessed data, and dynamically balances between these modes based on workload requirements. This single system replaces the need for separate dedicated fast memory and slow storage infrastructures
2Quantity of substance
If data is stored in remote memory pools, then storage capacity increases, but data transfer latency increases for time-sensitive applications
Solution Approach 1:
The patent implements predictive pre-fetching mechanisms that anticipate future memory access patterns. By analyzing historical access patterns and workload characteristics, the system proactively loads predicted needed data from remote memory pools into local memory before actual access occurs, thereby eliminating latency during critical access operations
Solution Approach 2:
The patent creates different memory quality zones: local memory provides high-speed access for frequently accessed data, while remote memory pools provide capacity for less frequently accessed data. The system intelligently places data in the appropriate zone based on access patterns, ensuring that time-sensitive operations always access data from the fast local zone while maintaining overall storage capacity through the remote zone
3Speed
If memory allocation is optimized for speed, then access latency decreases, but memory resource utilization decreases when demand is low
Solution Approach 1:
The patent implements dynamic memory allocation policies that adjust the balance between speed and utilization based on current workload conditions. During high-demand periods, the system prioritizes keeping sufficient data in fast local memory. During low-demand periods, it safely migrates data to remote memory pools, thereby maximizing resource utilization without permanently sacrificing access speed when needed
Data Source
AI summary
Examples described herein relate to providing an interface to an operating system (OS) to create different memory pool classes to allocate to one or more processes and allocate a memory pool class with a process of the one or more processes. In some examples, a memory pool class of the different memory pool classes defines a mixture of memory devices in at least one memory pool available for access by the one or more processes. In some examples, memory devices are associated with multiple memory pool classes to provide multiple different categories of memory resource capabilities.