Boost Java App Performance with MyPerf4J: A Complete Monitoring Guide

Feature Highlights

High performance: single‑thread, up to 16 million ops/sec, 63 ns per record.

Non‑intrusive: uses JavaAgent, no code changes required.

Real‑time: second‑level statistics with full‑volume collection.

Metrics: JVM metrics (Thread, Memory, GC, Class, etc.) and method metrics (RPS, latency percentiles, etc.).

Installation & Deployment

Download and unzip

MyPerf4J-ASM.zip

, then edit

MyPerf4J.properties

to set

app_name

,

metrics.log.xxx

and

filter.packages.include

.

Add the following JVM arguments:

<code>-javaagent:D:\MyPerf4J-ASM-3.3.0-SNAPSHOT\MyPerf4J-ASM-3.3.0-SNAPSHOT.jar
-DMyPerf4JPropFile=D:\MyPerf4J-ASM-3.3.0-SNAPSHOT\MyPerf4J.properties
</code>

Install InfluxDB v2.x, create a bucket named

MyPerf4J

, and configure the exporter:

<code>metrics.exporter = http.influxdb
influxdb.version=2.1
influxdb.orgName=org   # adjust to your org
influxdb.host=127.0.0.1
influxdb.port=8086     # adjust if needed
influxdb.username=test
influxdb.password=12345678
influxdb.database=MyPerf4J   # bucket name
</code>

Install Grafana (e.g., version 8.4.0) and import the Method Dashboard and JVM Dashboard.

In Grafana set the data source name to

InfluxDB

and use Flux as the query language. Example Flux query to list tag values:

<code># Application
import "influxdata/influxdb/schema"

schema.tagValues(
  bucket: "MyPerf4J",
  tag: "AppName",
  predicate: (r) => true,
  start: -7d
)

// Host
schema.tagValues(
  bucket: "MyPerf4J",
  tag: "host",
  predicate: (r) => true,
  start: -7d
)
</code>

Monitoring Effect

Project Address

<code>https://github.com/LinShunKang/MyPerf4J
</code>

Conclusion

MyPerf4J is a high‑concurrency, low‑latency Java performance monitoring and statistics tool. It supports log files and InfluxDB storage, works well with Grafana for real‑time visualization, and is worth trying for anyone interested in server monitoring.