API Usage Logging with Granular Observability and Bug Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current logging methods for API libraries on client devices are basic and lack the capability to provide detailed analytics, making it difficult to determine if an API should be deprecated or enhanced, and do not facilitate proactive bug hunting or efficient data processing.
Innovation Solution
A specialized data logging feature on client devices that captures detailed API usage metrics and observability, including error/failure states, debug states, and usage analytics, with configurable logging to separate internal and external developer groups, and supports plug-in architecture for event processing and enrichment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If basic logging methods are used for API libraries, then implementation simplicity is maintained, but analytics capability and bug detection ability are insufficient
Solution Approach 1:
The patent introduces an intermediary logging system that sits between the API library and the output destination. This intermediary captures detailed API usage events, enriches them with contextual information, and routes them to appropriate destinations. The intermediary handles the complexity of data collection and processing internally, while presenting a simple interface to developers through configuration files and standard logging mechanisms.
Solution Approach 2:
The logging system is segmented into distinct components: event capture modules for different API functions, data enrichment modules, filtering modules, and output modules. Each segment handles specific aspects of logging independently, allowing the system to achieve high analytics capability through modular design while maintaining ease of implementation through independent configuration of each segment.
2Loss of information
If detailed API usage data is captured and logged, then analytics capability improves, but data processing overhead and storage requirements increase
Solution Approach 1:
The system implements partial action by allowing selective logging of API events based on configuration. Users can specify which API functions and event types to log in detail, while other events are logged at summary level or not logged at all. This reduces data processing overhead and storage requirements while maintaining complete information for critical API usage patterns that require detailed analytics.
Solution Approach 2:
The logging system changes parameters dynamically based on runtime conditions. Log verbosity, buffer sizes, and processing intensity are adjusted according to device state, API call frequency, and configured priorities. This allows the system to maintain complete API usage information while optimizing data processing overhead by adapting to actual usage patterns rather than consistently processing all data at maximum detail.
3Reliability
If logging data is processed in real-time with high detail, then bug detection capability improves, but device performance and processing speed may be impacted
Solution Approach 1:
The system uses periodic action by implementing batch processing intervals for log data. Instead of processing every log event immediately, the system accumulates events in buffers and processes them in periodic batches. This reduces the real-time processing overhead on device performance while maintaining reliable bug detection capability, as the periodic processing ensures comprehensive analysis of API usage patterns without constant interruption of the main application throughput.
Solution Approach 2:
The logging system implements self-service through asynchronous operation and automatic buffer management. The logging subsystem serves itself by managing its own data buffers, processing queues, and resource allocation without requiring active intervention or blocking of the main application. This allows bug detection to proceed with high detail while minimizing impact on device productivity, as the logging system handles its own processing demands independently.
Data Source
AI summary
Systems, methods, and computer-readable storage mediums for performing specialized data logging on a client device are disclosed. Observability data pertaining to an API on a client device is captured. The observability includes, for each of one or more functions of the API, a number of times the function is used and a number of times each parameter of the function is used. The observability data pertaining to the API is communicated to a target based on a type of the observability data.