FunctionGraph Monitoring

This section describes the types, names, and meanings of FunctionGraph metrics collected by APM.

**Table 1** Collection parameters for FunctionGraph monitoring
Parameter	Data Type	Application Type	Default	Supported Start Agent Version	Supported End Agent Version	Description
Sampling Type	radio	JAVA	4	1.0.0	-	Sampling type. Options: full sampling, percentage sampling, fixed-quantity sampling per minute, and intelligent sampling (default).
Sampling Ratio	integer	JAVA	10	1.0.0	-	Percentage of samples to the total number of trace data records
Samples/Minute	integer	JAVA	1000	1.0.0	-	Number of trace data records collected every minute.
Slow Request Threshold	integer	JAVA	800	2.0.0	-	Slow request threshold. If the threshold is crossed, the method is regarded as a slow method. In that case, the trace sampling ratio will be increased by default.
Method Configuration	obj_array	JAVA	-	2.0.0	-	Configure the slow request threshold and sampling ratio for each method separately. The following sampling policies can be set: percentage sampling, fixed-quantity sampling per minute, and automatic sampling.

**Table 2** FunctionGraph metrics
Category	Metric	Name	Description	Unit	Data Type	Default Aggregation Mode
Method monitoring (detail: APM counts URL call statistics by function method.)	method	Method	Request method	-	ENUM	LAST
	concurrentMax	Max. Concurrency	Maximum concurrency of the method	-	INT	MAX
	errorCount	Errors	Number of times that the method fails to be called	-	INT	SUM
	invokeCount	Calls	Number of times that the method is called	-	INT	SUM
	maxTime	Max. RT	Maximum response time of the method in a collection period	ms	INT	MAX
	runningCount	Ongoing Executions	Number of executions of the method at the time of collection	-	INT	SUM
	range1	0–10 ms	Number of requests with 0–10 ms response time	-	INT	SUM
	range2	10–100 ms	Number of requests with 10–100 ms response time	-	INT	SUM
	range3	100–500 ms	Number of requests with 100–500 ms response time	-	INT	SUM
	range4	500–1000 ms	Number of requests with 500–1000 ms response time	-	INT	SUM
	range5	1–10s	Number of requests with 1–10s response time	-	INT	SUM
	range6	> 10s	Number of requests with response time longer than 10s	-	INT	SUM
	totalTime	Total RT	Total response time of the method	-	INT	SUM
Cluster call (cluster: APM counts URL call statistics based on the cluster ID of the caller.)	clusterId	Cluster ID	Cluster ID of the caller	-	ENUM	LAST
	invokeCount	Calls	Number of times the cluster is called	-	INT	SUM
	totalTime	Total RT	Total response time for calling the cluster	ms	INT	SUM
	errorCount	Errors	Number of times that the cluster fails to be called	-	INT	SUM