hotspotmethod

In the field of application performance monitoring and optimization, Hotspot Methods are not independent tools, but refer to "core functions/interfaces that are frequently called, consume large amounts of system resources (CPU/memory/IO), or take too long to execute". Their core "function" lies in serving as a key entry point for performance analysis — by identifying, locating, and optimizing hotspot methods, system performance bottlenecks can be efficiently resolved, resource consumption reduced, and business stability ensured.

Preconditions

1. The continuous profiling function needs to be enabled during probe installation. For specific enabling operations, please refer to the relevant documentation.
2. Before using the hotspot methods function, you need to enable the hotspot method identification configuration first. For specific operations, please refer to Entity Identification -> Hotspot Methods.

Usage Scenarios

• Performance Assurance for High-Concurrency Businesses: Locate core methods that are "frequently called + time-consuming", such as the "inventory deduction method" in flash sale scenarios (call frequency of 5,000 times/second, single execution time of 80ms);
• Troubleshooting of Intermittent/Periodic Faults: Through continuous monitoring, capture temporary hotspots during fault periods. For example, "at 19:00, the dynamic loading method’s call frequency surged from 100 times/second to 2,000 times/second due to a sharp increase in user-published content, and the internal sensitive word filtering logic was inefficient".

Get Started

1. Navigate to the System Services -> Services -> Service Details -> Analysis page, where you can find the entry to Hotspot Methods.
   

2. The Hotspot Methods Details page supports statistics of Stacktrace samples, trend statistics, and display of code stack data.
   

3. For Hotspot Methods Top statistics, switch to the Hotspot Methods Top tab to view the high-frequency methods in the current analysis.