OS Session Failover via VM Mapping and Distributed Cache
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In multi-tier architectures, errors at the Operating System (OS) level are difficult to recover, leading to data loss, necessitating effective session failover methods at the OS level.
Innovation Solution
The method involves selecting a Virtual Machine (VM) OS to perform failover when a server OS fails, mapping its memory and storage to a distributed share-memory-and-cloud-storage system via a hypervisor, and obtaining the OS state from a distributed cache system to ensure seamless data access and application continuity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional server OS is used without failover mechanism, then system simplicity is maintained, but reliability deteriorates due to inability to recover from OS level errors
Solution Approach 1:
The patent implements preliminary action by maintaining a standby Virtual Machine OS that is pre-configured and ready to take over the failed server OS. The health monitor continuously checks the status of the server OS, and when failure is detected, the standby VM is immediately activated to perform failover, ensuring continuous operation without interruption.
Solution Approach 2:
The patent introduces a health monitor as an intermediary component that monitors the server OS status and triggers failover when necessary. The health monitor acts as a mediator between the server OS and the standby VM, coordinating the failover process and ensuring smooth transition without direct intervention in the server OS operations.
2Reliability
If VM OS failover mechanism is implemented, then reliability is improved through automatic failover, but device complexity increases due to additional components
Solution Approach 1:
The patent applies copying by creating a standby Virtual Machine OS that is a copy of the server OS. This standby VM includes identical hardware configurations, software applications, and data structures, allowing it to seamlessly take over the failed server's functions. The copying approach enables failover while maintaining system functionality.
Solution Approach 2:
The standby Virtual Machine OS is designed with multi-functionality to serve multiple purposes: it acts as a backup system, a testing environment, and a ready-to-takeover replica. This universal design allows a single VM to perform multiple functions, reducing the need for separate dedicated failover infrastructure.
3Ease of operation
If distributed share-memory-and-cloud-storage is used, then data accessibility is improved during failover, but system complexity increases due to mapping and integration requirements
Solution Approach 1:
The hypervisor serves as an intermediary between the Virtual Machine OS and the distributed share-memory-and-cloud-storage system. It automatically maps the VM's memory and storage to the distributed storage infrastructure, handling the complexity of resource allocation and data access without requiring manual configuration from the VM or application layer.
Solution Approach 2:
The system implements self-service by allowing the Virtual Machine OS to automatically access and map distributed storage resources without human intervention. The hypervisor and distributed storage system work together to transparently provide file access, memory mapping, and data persistence services to the VM, eliminating the need for complex manual setup.
Data Source
AI summary
The invention introduces a method for session failover in OS (Operating System) level, which contains at least the following steps. A VM (Virtual Machine) OS is selected to perform a failover when a server OS is determined to have failed. A memory and a local storage device of the VM OS are mapped to the physical memory space and physical storage space of a distributed share-memory-and-cloud-storage, which were allocated for the failed server OS via a hypervisor corresponding to the VM OS. The OS state of the failed server OS is obtained from a distributed cache system via the hypervisor.