Cloud API Operation Management Through User-Service Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies fail to accurately determine which users are affected by failures in cloud-native applications, particularly when failures occur in services provided by third parties, as they do not consider the specific functions used by individual users.
Innovation Solution
An operation management apparatus that manages system operations via APIs, using test data to detect failures in call destination services and confirm their reproduction, thereby identifying affected users through a confirmation test process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If failure detection is performed for all users, then reliability of failure notification is improved, but device complexity increases due to need to track user-service-API relationships
Solution Approach 1:
The system pre-establishes and stores the correspondence relationships between users, services, and APIs before failures occur. This preliminary action allows the system to quickly determine affected users when a failure is detected, without needing to analyze complex relationships in real-time during failure events.
Solution Approach 2:
The system creates and maintains a copied representation of the user-service-API relationship structure in the operation management apparatus. This copied structure enables efficient querying and analysis of affected users without directly accessing or modifying the actual service systems, reducing complexity while maintaining reliability.
2Measurement precision
If confirmation tests are executed for all services, then measurement precision of failure impact is improved, but productivity decreases due to additional test execution time
Solution Approach 1:
The system applies confirmation tests selectively based on local conditions - specifically, tests are executed only for services that have established correspondence relationships with users and are identified as potentially affected by failures. This localized approach maintains measurement precision for relevant services while avoiding unnecessary tests on unrelated services.
Solution Approach 2:
The system performs partial confirmation tests only on the subset of services that are actually affected by failures and have user correspondences, rather than executing tests on all services. This partial action approach achieves sufficient measurement precision for failure impact assessment while significantly reducing the time and resources required compared to comprehensive testing.
3Loss of information
If user-specific service usage information is tracked, then loss of information about affected users is reduced, but device complexity increases due to additional data storage requirements
Solution Approach 1:
The correspondence relationship data structure serves multiple functions simultaneously: it tracks which users are affected by service failures, enables targeted notification, and provides the basis for confirmation testing. This multi-functionality reduces the need for separate data structures and minimizes overall data management complexity while preventing information loss.
Data Source
AI summary
In an operation management apparatus that manages an operation of a system that provides a service requesting processing from an external service via an application programming interface (API), a storage device stores test data for causing the external service to execute processing via the API, the external service that a user uses through the service, and a correspondence relationship between the external service and the API that the user uses the processor is configured to detect whether or not a failure has occurred on the external service side, confirm whether or not the failure on the call destination service side is reproduced via an API used when the failure occurred, using the test data in a case of detecting the failure and specify a user who uses the external service and the API used, in a case where the failure on the external service side is reproduced.