Application Group Checkpointing via File System Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current enterprise and wireless systems lack mechanisms for protecting multi-application services, which are composed of individual applications from different vendors and operate at different times, leading to challenges in ensuring high availability and data integrity across distributed storage systems.
Innovation Solution
A method and system for checkpointing storage related to individual applications, allowing transparent migration and fail-over of application groups while maintaining client connection transparency, using a coordinator to manage registration, fault detection, and recovery across nodes, and supporting both fork() and exec() operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If storage checkpointing is implemented at the block level of the storage subsystem, then storage protection is achieved, but all applications on the disk are affected including unrelated applications
Solution Approach 1:
The patent segments storage checkpointing from the block level to the file level, creating independent checkpoint groups for different applications. Each application or process is assigned its own checkpoint group, allowing selective restoration of individual applications without affecting others on the same storage device. This segmentation enables granular control over which applications are restored when a checkpoint is recovered.
Solution Approach 2:
The patent implements local quality by allowing different checkpoint policies and restoration behaviors for different applications. Each application can have its own checkpoint group with customized parameters, enabling tailored protection strategies for critical versus non-critical applications while maintaining system-wide storage management.
2Reliability
If customized applications run on custom hardware to achieve high availability, then availability is improved, but cost and complexity increase
Solution Approach 1:
The patent creates a universal checkpointing system that works across multiple operating systems and hardware platforms. The checkpoint group mechanism provides high availability functionality through software rather than requiring custom hardware, making the solution portable and applicable to diverse computing environments including Windows, Linux, and other OSes.
Solution Approach 2:
The patent uses copying of file system state to create checkpoints, storing checkpoint data in designated directories or external storage locations. This copying approach enables restoration capabilities without requiring specialized hardware, using standard file copying and restoration techniques that work across conventional computing platforms.
3Reliability
If multiple independent applications are checkpointed together, then comprehensive protection is achieved, but restoration complexity increases
Solution Approach 1:
The patent segments applications into distinct checkpoint groups, allowing administrators to organize applications by function, criticality, or ownership. This segmentation simplifies restoration management by enabling selective restoration of specific groups rather than requiring complex coordination of all applications individually.
Solution Approach 2:
The patent introduces checkpoint groups as an intermediary layer between the file system and individual applications. These groups act as management units that simplify the restoration process by providing a single point of control for restoring multiple related applications, reducing the complexity of coordinating restorations across numerous independent applications.
Data Source
AI summary
A system and method for storage checkpointing to a group of independent computer applications. The system has a storage disk that stores files; a storage access interface to access the storage disk; and a computer. The computer runs the group of independent computer applications and utilizes the files stored on the storage disk. A file system on the server accesses the files stored on the storage disk. An operating system and at least one device driver can be called by the file system, and at least one buffer buffers first data written to the storage disk and second data read from the storage disk.