Application-Aware Memory Patrol Scrubbing for Bandwidth Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current memory error handling in cloud deployments, particularly in large-scale SAP HANA and hyperscaler environments, is inefficient due to the lack of targeted and adaptive memory scrubbing mechanisms, leading to potential data loss and performance degradation.
Innovation Solution
The implementation of an adaptive patrol scrubber that focuses on critical memory regions specified by the OS and/or application software, incorporating application-specific checksumming and correction logic, and integrating Quality of Service (QoS) APIs to prioritize memory reliability alongside bandwidth and cache capacity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional patrol scrubbing periodically cycles through all memory using ECC logic, then single-bit errors are corrected before accumulation, but multi-bit errors cannot be addressed and critical memory regions receive equal treatment wasting bandwidth
Solution Approach 1:
The patent implements application-aware patrol scrubbing that identifies and prioritizes critical memory regions based on application-specific requirements. The system uses application hints to determine which memory regions are most important and directs scrubbing resources accordingly, applying different scrubbing frequencies and intensities to different memory regions rather than treating all memory uniformly.
Solution Approach 2:
The memory space is segmented into critical and non-critical regions based on application requirements. The patrol scrubbing mechanism is divided into targeted scrubbing for critical regions and reduced or skipped scrubbing for non-critical regions, allowing bandwidth to be allocated proportionally to actual need.
2Loss of energy
If application-aware patrol scrubbing targets only critical memory regions, then bandwidth consumption is reduced, but coverage of non-critical regions may be insufficient
Solution Approach 1:
The system performs partial scrubbing by focusing intensive scrubbing efforts on critical memory regions that require high reliability, while applying minimal or no scrubbing to non-critical regions where errors are acceptable. This partial action approach optimizes the balance between reliability and resource consumption by doing just enough scrubbing where it matters most.
3Productivity
If hardware ECC logic is used for error correction, then hardware correctable errors are fixed with no software involvement, but software-based error detection and correction capabilities are not utilized
Solution Approach 1:
The patent merges hardware ECC capabilities with software-based error detection and correction mechanisms. The system combines hardware patrol scrubbing with software application-aware scrubbing, creating a hybrid approach that leverages both hardware efficiency and software flexibility to provide comprehensive error handling.
Data Source
AI summary
Methods and apparatus for application aware memory patrol scrubbing techniques. The method may be performed on a computing system including one or more memory devices and running multiple applications with associated processes. The computer system may be implemented in a multi-tenant environment, where virtual instances of physical resources provided by the system are allocated to separate tenants, such as through virtualization schemes employing virtual machines or containers. Quality of Service (QoS) scrubbing logic and novel interfaces are provided to enable memory scrubbing QoS policies to be applied at the tenant, application, and/or process level. This QoS policies may include memory ranges for which specific policies are applied, as well as bandwidth allocations for performing scrubbing operations. A pattern generator is also provided for generating scrubbing patterns based on observed or predicted memory access patterns and/or predefined patterns.