Open Platform Memory Management via Remote Service Offloading
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Open platform systems face instability and resource overloading due to insufficient memory resources, leading to out-of-memory exceptions and service interruptions, which existing technologies fail to efficiently address.
Innovation Solution
A memory management system that dynamically monitors CPU and memory usage, searches for remote open platforms over the internet, and offloads memory-intensive tasks to a remote platform when local memory thresholds are exceeded, using a CPU monitoring bundle, memory monitoring bundle, discovering bundle, management agent bundle, and proxy bundle to access standard service instances.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If service applications are dynamically added to the open platform, then service diversity is improved, but memory resource consumption increases leading to resource overloading
Solution Approach 1:
The patent segments the memory management function into multiple monitoring bundles (CPU monitoring bundle, memory monitoring bundle) that independently monitor different resources. When memory threshold is exceeded, the system segments service instances into local instances and remote instances, offloading memory-intensive services to remote platforms through the service registry, thereby resolving the contradiction between service diversity and memory consumption
Solution Approach 2:
The patent introduces a new dimension of remote platform architecture to solve local memory constraints. By extending the service execution environment from purely local to include remote platforms accessible via internet/intranet, the system achieves unlimited service diversity without being constrained by local memory resources
2Productivity
If memory space is allocated for service execution, then service functionality is improved, but system stability deteriorates when out of memory exceptions occur
Solution Approach 1:
The patent implements preliminary action by setting memory usage thresholds and establishing monitoring mechanisms before memory exhaustion occurs. The management agent bundle proactively identifies when memory usage exceeds the threshold and triggers service offloading to remote platforms, preventing out-of-memory exceptions and maintaining system stability
Solution Approach 2:
The patent introduces the service registry and proxy bundle as intermediaries between local services and remote platforms. When memory pressure is detected, the proxy bundle acts as a mediator to redirect service instance execution to remote platforms, ensuring smooth transition without system crashes and maintaining continuous service functionality
3Quantity of substance
If memory monitoring and management mechanisms are added, then memory resource management is improved, but device complexity increases
Solution Approach 1:
The patent implements universal bundles that perform multiple functions: the CPU monitoring bundle monitors both CPU usage and triggers memory management decisions, while the service registry serves both as a service discovery mechanism and as a routing table for local/remote service instances. This multi-functionality reduces the need for separate dedicated components, managing complexity
4Quantity of substance
If service instances are suspended due to out of memory exceptions, then local memory usage is reduced, but service continuity deteriorates
Solution Approach 1:
The patent creates remote copies of service instances on distant platforms. When local memory is insufficient, the proxy bundle redirects service execution to these remote copies, which have access to different memory resources. This copying approach maintains service continuity while managing local memory usage
Data Source
AI summary
The present invention relates to a memory management system and method for open platform. The memory management system and method of the present invention utilizes the main spirit of sharing service in open platform. When the used memory in local open platform exceeds an upper limit, the standard service bundle access interface is used for accessing the standard service bundle in remote open platform. Therefore, the standard service bundle in local open platform can be off-loaded to release the memory space so as to resolve the memory shortage problem. The stability of the whole system can be maintained.