Persistent Memory Modules with Secondary Network Access Path
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current memory storage systems, particularly DRAM and flash-backed DRAM, face challenges in preserving data during power failures and providing network-based access to retained data, leading to data unavailability and delayed service resumption.
Innovation Solution
A persistent memory system with a secondary access path and network-based access via Ethernet, enabling authorized devices to access retained data even after a power failure, using a Power-over-Ethernet switch to supply power to memory modules and controllers, allowing data extraction over a communications network.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single interface is used for persistent memory modules, then device complexity is reduced, but data availability deteriorates when the hosting server fails
Solution Approach 1:
The patent divides the access path into two independent segments: a primary interface for normal operation and a secondary network interface for backup access. This segmentation allows the system to maintain simple primary operation while adding redundancy through a separate access path that can be used when the primary path fails.
Solution Approach 2:
The patent introduces a network interface card (NIC) as an intermediary component that provides an alternative access path to the persistent memory module. This mediator enables remote systems to access retained data through a network connection when the primary hosting server is unavailable, without complicating the primary memory interface.
2Duration of action of stationary object
If Flash-backed DRAM is used to retain data after power failure, then data retention is improved, but data accessibility deteriorates when the host server is inactive
Solution Approach 1:
The patent implements preliminary action by pre-configuring a secondary network interface and establishing an alternative access path before the primary server fails. The NIC is预先 equipped with the capability to access the persistent memory module directly, so when the host server becomes inactive, the data is immediately accessible through the pre-established network path without requiring any recovery or reconfiguration actions.
Solution Approach 2:
The network interface card serves as an intermediary that bridges the persistent memory module and remote systems. This mediator allows data to be accessed through the network infrastructure when the primary hosting server is unavailable, effectively decoupling data accessibility from the status of the original host server.
3Use of energy by moving object
If Power-over-Ethernet is used to supply power to memory modules during power failure, then energy availability is improved, but device complexity increases
Solution Approach 1:
The patent applies universality by using the Ethernet cable to serve multiple functions: both data communication and power delivery. The existing network infrastructure is leveraged to provide power to the persistent memory module and NIC during power failures, eliminating the need for separate power supply circuits and reducing overall system complexity despite the dual functionality requirement.
Solution Approach 2:
The system implements self-service by using the Power-over-Ethernet capability to automatically supply power to the memory module and network interface when external power is unavailable. The NIC detects the power failure condition and utilizes the PoE infrastructure to maintain operational status without requiring external intervention or complex power management logic.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Ensures data accessibility and rapid service resumption by providing network-based access to retained memory, even during power failures, and allows authorized devices to extract data from other servers, enhancing data availability and system resilience.
Implementation Method 1
using a Power-over-Ethernet switch to supply power to memory modules and controllers
Data Source
AI summary
A persistent memory storage system and method of operating. The storage system is embodied as memory module (e.g., a DIMM or PCIe card) in a host server and accessible via a first access path to/from the memory module. After a power failure at the host server, any authorized device, e.g., a second server, is configured to access and extract the retained memory at the failed host server via a secondary network-based access path (e.g., Ethernet connection). Power may be supplied to the memory module of the failed host server via a Power-over-Ethernet switch to support the secondary access path. A Power-over-Ethernet (PoE) switch may provide the necessary power to the memory module when power is unavailable from the host server and when any temporary power source is exhausted. A redundant power source and/or backup battery may supply the PoE switch. Such power sources and/or batteries can thereby provide backup power.