README-EN.md

Profiler

English / 中文

Introduction

Profiler is a continuous profiling tool that based on go pprof and go trace

• Supported Sample
  • trace fgprof profile mutex heap goroutine allocs block threadcreate
• Hot reloading configuration
  • Collect samples of the target service according to the configuration file
  • The collection program will watch the changes of the configuration file and apply the changed immediately
• Chart Trend
  • Provide charts to observe the trend of multiple service performance indicators and find the time point of performance problems
  • Each bubble is a sample file of Profile and Trace
• Detailed Analysis
  • Click the bubbles in the charts to jump to the detailed page of Profile and Trace for further detailed analysis

Demo

Chart trend	Click the bubble to jump the detailed profile
Click the bubble to jump to the detailed trace	Click the bubble to jump to the detailed trance

Getting Started

Local

Run server on port 8080

go run server/main.go 

Run ui on port 80

cd ui
npm install --registry=https://registry.npm.taobao.org
npm run dev --base_api_url=http://localhost:8080

Docker

docker run -d -p 80:80 --name profiler xyctruth/profiler:latest

Using custom configuration file

mkdir ~/profiler-config/
cp ./collector.yaml ~/profiler-config/
docker run -d -p 80:80 -v ~/profiler-config/:/profiler/config/ --name profiler xyctruth/profiler:latest

Using persistent data

docker run -d -p 80:80 -v ~/profiler-data/:/profiler/data/ --name profiler xyctruth/profiler:latest

Helm

Install the Profiler chart:

helm install --create-namespace -n profiler-system profiler ./charts/profiler

More on Helm docs

Collector configuration

The golang program that needs to be collected and analyzed needs to provide the net/http/pprof endpoint and configure it in the ./collector.yaml configuration file.

The configuration file can be updated online, and the collection program will monitor the change of the configuration file and apply the changed configuration file immediately.

collector.yaml

collector:
  targetConfigs:

    profiler-server:        # Target name
      interval: 15s         # Scrape interval
      expiration: 0         # No expiration time
      instances: ["localhost:9000"]  # Target service host
      labels:
        namespace: f005
        type: gateway
      profileConfigs:       # Use default configuration

    server2:
      interval: 10s
      expiration: 168h      # Expiration time seven days
      instances: ["localhost:9000"]
      labels:
        namespace: f004
        type: svc
      profileConfigs:       # Override some default configuration fields
        trace:
          enable: false
        fgprof:
          enable: false
        profile:
          path: /debug/pprof/profile?seconds=10
          enable: false
        heap:
          path: /debug/pprof/heap

default configuration of profileConfigs

The default trace analysis is turned off, because the trace file is too large, about (500KB ~ 2M), you need to open the trace analysis in the collector.yaml setting to override the default trace configuration.

profileConfigs:
  profile:
    path: /debug/pprof/profile?seconds=10
    enable: true
  fgprof:
    path: /debug/fgprof?seconds=10
    enable: true
  mutex:
    path: /debug/pprof/mutex
    enable: true
  heap:
    path: /debug/pprof/heap
    enable: true
  goroutine:
    path: /debug/pprof/goroutine
    enable: true
  allocs:
    path: /debug/pprof/allocs
    enable: true
  block:
    path: /debug/pprof/block
    enable: true
  threadcreate:
    path: /debug/pprof/threadcreate
    enable: true
  trace:
    path: /debug/pprof/trace?seconds=10
    enable: false

JetBrains OSS License