Application Instance State Replication for Multi-Tenancy Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multi-tenancy architectures in cloud computing face inefficiencies in resource utilization and storage space due to coarse granularity of sharing among multiple instances of applications, leading to unnecessary resource consumption and limited space saving, especially in virtual machine-level replication.
Innovation Solution
A method and system for running multiple instances of an application by starting a root instance, stopping its threads at a preset freezing point, and replicating the running state of these threads to additional instances, allowing for more efficient data sharing and reduced storage space occupation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If virtual machine-level replication is used for multi-tenancy, then application instances can be shared among multiple tenants, but storage space consumption increases due to coarse granularity of sharing
Solution Approach 1:
The patent segments the application instance into two parts: a shared root instance (common code and resources) and individual thread instances (specific execution states). This segmentation allows the root instance to be shared among multiple tenants while only replicating the necessary thread-level data, thereby reducing storage space consumption while maintaining multi-tenancy capability.
Solution Approach 2:
The patent implements a nested structure where multiple thread instances are nested within a single shared root instance. The root instance contains the common code and resources, while individual thread instances contain only their specific execution states. This nested arrangement enables efficient space utilization by avoiding redundant storage of common data across multiple instances.
2Reliability
If multiple full instances of an application are run for each tenant, then each tenant has dedicated resources, but resource efficiency decreases due to repeated code loading
Solution Approach 1:
The patent merges multiple tenant instances into a single shared root instance while maintaining separate thread-level execution states. This combining approach allows multiple tenants to share the same code base and common resources in the root instance, eliminating repeated code loading and improving resource efficiency, while still providing each tenant with isolated thread instances for their specific operations.
Data Source
AI summary
Running of a root instance of an application is started. The root instance includes at least one thread. In response to determining that a thread of the root instance runs to a preset freezing point in the application, running of all threads of the root instance is stopped. In response to starting to run an additional instance of the application, a running state of all threads of the root instance is replicated as a running state of all threads of the additional instance of the application. Running all threads of the additional instance of the application is continued.


