Cloud Service Manager for Medical Application Scalability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Medical imaging applications in public cloud data centers face scalability issues due to the need for repeated loading of extensive medical data, which is time-consuming and renders the application inoperable, as conventional systems do not support fixed instance allocation to users.
Innovation Solution
Implementing a service manager instantiation for each user upon initial request, which maintains a single application instance throughout the user session, ensuring data is loaded only once and reducing the need for repeated loading.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional cloud systems use dynamic provisioning with Round Robin allocation, then scalability is improved, but application responsiveness deteriorates due to repeated data loading
Solution Approach 1:
The service manager performs preliminary actions by pre-loading medical image data into memory before the user actually needs it. When a user requests an application instance, the service manager proactively loads the required medical data (DICOM images, patient information, study data) into the application instance's memory space in advance, so that when the user interacts with the application, the data is already available and no time is lost during the session.
Solution Approach 2:
The service manager acts as an intermediary between the cloud management system and the application instances. It sits in the middle layer, receiving requests from users, managing the lifecycle of application instances, pre-loading data into these instances, and coordinating with the cloud management system for instance allocation. This intermediary role allows it to optimize data loading timing and maintain persistent instances without requiring changes to the underlying cloud infrastructure.
2Device complexity
If medical image data is loaded for each user request, then application instance allocation is simplified, but user experience deteriorates due to noticeable loading time
Solution Approach 1:
The service manager performs preliminary data loading into application instance memory before the user needs the data. When allocating an instance to a user, the service manager has already loaded the medical image data, patient information, and study data into the instance's memory space. This preliminary action ensures that when the user accesses the application, there is no noticeable loading time, and the application responds immediately to user interactions throughout the session.
3Use of energy by moving object
If application instances are stateless and dynamically provisioned, then resource utilization is improved, but data availability deteriorates for multi-request user sessions
Solution Approach 1:
The service manager introduces dynamic instance management where application instances transition between different states: idle (pre-loaded with data), active (assigned to user), and terminated (when user session ends). The service manager dynamically adjusts the lifecycle of instances based on user session duration and activity patterns, maintaining instances in memory during active sessions to ensure data availability, while allowing the underlying cloud infrastructure to dynamically provision and deprovision instances based on overall system load and resource utilization.
Data Source
AI summary
A method for providing scaled applications in a cloud data center is disclosed. The method includes receiving a request from a user to use one of a plurality of applications maintained in a memory of a cloud data center. The method further includes assigning one of a plurality of instantiations of a service manager to the user and establishing, using the assigned instantiation of the service manager, a connection between the user and an instance of the requested application for a duration of a user session.