Microservice Orchestration for Multi-Cloud Execution Speed
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In multi-cloud environments, the frequent communication between distributed bases for data analysis leads to slow execution speeds due to increased trial and error cycles, as each microservice communication occurs over wide-area networks.
Innovation Solution
An information processing system with a management device that coordinates microservices and data movement between distributed bases based on performance information, optimizing the execution order and location of microservices and data to reduce unnecessary communications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If microservices are executed in distributed bases via WAN communication, then data analysis processing can be performed across multiple clouds, but the execution speed becomes slow due to frequent communications
Solution Approach 1:
The patent applies preliminary action by pre-fetching data from distributed bases before the actual analysis processing begins. The management device predicts which data will be needed and transfers it in advance, so that when microservices need to access the data, it is already available locally, eliminating WAN communication delays during execution.
Solution Approach 2:
The patent introduces an intermediary mechanism in the form of a management device that coordinates between distributed bases and microservices. This intermediary manages data placement and movement strategically, acting as a buffer that reduces direct WAN communications between microservices and distributed bases, thereby improving execution speed while maintaining multi-cloud adaptability.
2Measurement precision
If trial and error cycles are run frequently to improve data analysis accuracy, then analysis accuracy improves, but the execution speed of the trial and error cycle slows down due to increased communications
Solution Approach 1:
The patent applies preliminary action by pre-positioning datasets in distributed bases before trial and error cycles begin. When multiple trial runs are needed for accuracy improvement, the data is already available locally in the distributed bases, allowing rapid iteration without repeated WAN communications, thus maintaining both accuracy and productivity.
Solution Approach 2:
The patent changes the parameter of data location from centralized to distributed across multiple bases. By strategically placing data copies in distributed bases based on access patterns, the system enables fast local access during trial and error cycles, improving execution speed while maintaining data analysis accuracy through access to the same datasets.
3Ease of operation
If data is scattered in distributed bases, then data can be stored closer to processing needs, but communication by WAN between bases occurs frequently and cycle execution speed becomes slow
Solution Approach 1:
The patent applies preliminary action by predicting future data access needs and pre-transferring data to distributed bases before actual access occurs. This reduces the frequency of WAN communications during execution while maintaining data accessibility, as data is moved proactively rather than reactively during trial and error cycles.
Solution Approach 2:
The patent uses copying by creating and distributing multiple copies of datasets across different bases based on predicted access patterns. Instead of moving the same physical data instance repeatedly via WAN, the system creates local copies that can be accessed quickly, reducing communication time while maintaining data accessibility from multiple locations.
Data Source
AI summary
An information processing system including Application Platform capable of communicating with Edge1 connected to each other to be able to communicate each other, in which Application Platform includes a second processor, information on microservices and data possessed by Edge1, and performance information describing the performance of Edge1, and the second processor uses predetermined data to combine a plurality of predetermined microservices and causes Edge1 to execute them in a predetermined order. When executing the application, microservices and data are moved between Edge1 based on the information of the microservices and the data possessed by Edge1, and the performance information.