Dynamic Distributed Tracing With Selective Synthetic Spans
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Distributed tracing in microservice architectures generates large amounts of tracing data, consuming system bandwidth and increasing latency, leading to inefficient debugging of underperformance issues due to manual code instrumentation being labor-intensive and often too late to address the problem.
Innovation Solution
Implement dynamic distributed tracing instrumentation with span generation components that monitor method calls and internal processes, identifying underperformance by comparing execution times to thresholds, and inserting synthetic spans to pinpoint the cause of performance issues.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If distributed tracing instrumentation is implemented to monitor all method calls and internal processes, then measurement precision of performance data is improved, but system bandwidth consumption increases and latency increases
Solution Approach 1:
The system performs preliminary actions by pre-defining performance thresholds and instrumentation policies before issues occur. When performance degradation is detected, the system has pre-configured synthetic span insertion mechanisms ready to execute immediately, enabling rapid response without full-blown tracing of all operations.
Solution Approach 2:
The patent extracts only the necessary tracing information by inserting synthetic spans at specific critical points rather than instrumenting all method calls. This selective extraction approach obtains sufficient performance data to identify bottlenecks while minimizing the volume of data transmitted and processed.
2Measurement precision
If distributed tracing instrumentation is implemented to monitor all method calls and internal processes, then measurement precision of performance data is improved, but system latency increases
Solution Approach 1:
The system performs preliminary actions by pre-defining performance thresholds and instrumentation policies before issues occur. When performance degradation is detected, the system has pre-configured synthetic span insertion mechanisms ready to execute immediately, enabling rapid response without full-blown tracing of all operations.
Solution Approach 2:
The patent extracts only the necessary tracing information by inserting synthetic spans at specific critical points rather than instrumenting all method calls. This selective extraction approach obtains sufficient performance data to identify bottlenecks while minimizing the volume of data transmitted and processed.
3Device complexity
If manual code instrumentation is used to identify underperformance causes, then device complexity is reduced, but productivity decreases due to labor-intensive and late debugging
Solution Approach 1:
The system implements self-service by automatically detecting performance degradation through monitored metrics and autonomously inserting synthetic spans to identify root causes. This eliminates the need for manual instrumentation while maintaining high debugging productivity through automated analysis and root cause identification.
Solution Approach 2:
The system continuously monitors performance metrics and provides feedback when thresholds are violated. This triggers automated synthetic span insertion and analysis, creating a closed-loop system that self-diagnoses performance issues without manual intervention and accelerates the debugging process.
Data Source
AI summary
A method call of a service to be analyzed is identified. A synthetic span is inserted to acquire timing information associated with a first set of source code of the method call. The timing information associated with the first set of source code acquired by the synthetic span is received.