`Arthas` is an Java Diagnostic tool open sourced by Alibaba.
Arthas can help developer trouble-shooting production issues for Java applications without modifying your code or restating your server.
`Arthas` 是Alibaba开源的Java诊断工具，深受开发者喜爱。
[中文说明](README_CN.md)
当你遇到以下类似问题而束手无策时，`Arthas`可以帮助你解决：
### Background
0. 这个类从哪个 jar 包加载的？为什么会报各种类相关的 Exception？
0. 我改的代码为什么没有执行到？难道是我没 commit？分支搞错了？
0. 遇到问题无法在线上 debug，难道只能通过加日志再重新发布吗？
0. 线上遇到某个用户的数据处理有问题，但线上同样无法 debug，线下无法重现！
0. 是否有一个全局视角来查看系统的运行状况？
0. 有什么办法可以监控到JVM的实时运行状态？
In production system, sometimes the networks is unreachable from local development environment. If you encounter some issues in production system, it is impossible to use IDE to debug the application remotely. More importantly, debugging in production is unacceptable, because it will suspend all the threads, which leads to business unavailability.
Some may think of reproducing the same issue on the test/staging environment, however, some tricky issue either can hardly be reproduced on a different environment, or even disappeared once restarted.
Thinking of adding some logs to your code? You have to go through test, staging, and then on to production. Time is money! That is a lot inefficient! Besides, the issue may not be reproducible once restart your JVM, as described above.
Arthas is born to solve these issues. You can trouble-shoot your production issue on-the-fly. No JVM restart, no additional code changes. Arthas works as an observer, which will never suspend your existing threads.
### Key features
### 快速开始
* Check whether a class is loaded? And whether is class loaded from? (Useful for trouble-shooting jar file conflicts)
* Decompile a class to ensure the code is running as expected.
* View classloader statistics, e.g. how may classloaders are there? how many class is loaded per classloader? What is the classloader hierarchy? Is there possible classloader leaks?
* View the method invocation details, e.g. method parameter, return object, thrown exception, and etc.
* Check the stack trace of specified method invocation. This is useful when you would like to know who is calling your method?
* Trace the method invocation to find slow sub-invocations.
* Monitor method invcation statistics, e.g. qps, rt, success rate and etc.
* Monitoring your system metrics, thread states and cpu usage, gc statistics, and etc.
* Support command line interactive mode, with auto completed feature enabled.
* Support telnet and websocket, which enables both local and remote diagnostics with command line and browsers.
### Quick start
#### Linux/Unix/Mac
安装Arthas:
Install Arthas:
```
curl -L https://alibaba.github.io/arthas/install.sh | sh
```
启动Arthas:
Start Arthas:
```
./as.sh
@ -40,25 +49,25 @@ curl -L https://alibaba.github.io/arthas/install.sh | sh
1. Click [here](http://search.maven.org/classic/#search%7Cga%7C1%7Cg%3A%22com.taobao.arthas%22%20AND%20a%3A%22arthas-packaging%22) to Download the latest binary releases.
`Arthas` is an Java Diagnostic tool open sourced by Alibaba.
Arthas can help developer trouble-shooting production issues for Java applications without modifying your code or restating your server.
English version goes [here](README.md).
### Background
`Arthas` 是Alibaba开源的Java诊断工具，深受开发者喜爱。
In production system, sometimes the networks is unreachable from local development environment. If you encounter some issues in production system, it is impossible to use IDE to debug the application remotely. More importantly, debugging in production is unacceptable, because it will suspend all the threads, which leads to business unavailability.
当你遇到以下类似问题而束手无策时，`Arthas`可以帮助你解决：
Some may think of reproducing the same issue on the test/staging environment, however, some tricky issue either can hardly be reproduced on a different environment, or even disappeared once restarted.
0. 这个类从哪个 jar 包加载的？为什么会报各种类相关的 Exception？
0. 我改的代码为什么没有执行到？难道是我没 commit？分支搞错了？
0. 遇到问题无法在线上 debug，难道只能通过加日志再重新发布吗？
0. 线上遇到某个用户的数据处理有问题，但线上同样无法 debug，线下无法重现！
0. 是否有一个全局视角来查看系统的运行状况？
0. 有什么办法可以监控到JVM的实时运行状态？
Thinking of adding some logs to your code? You have to go through test, staging, and then on to production. Time is money! That is a lot inefficient! Besides, the issue may not be reproducible once restart your JVM, as described above.
Arthas is born to solve these issues. You can trouble-shoot your production issue on-the-fly. No JVM restart, no additional code changes. Arthas works as an observer, which will never suspend your existing threads.
### Key features
* check whether a class is loaded? And whether is class loaded from? (Useful for trouble-shooting jar file conflicts)
* decompile a class to ensure the code is running as expected.
* view classloader statistics, e.g. how may classloaders are there? how many class is loaded per classloader? What is the classloader hierarchy? Is there possible classloader leaks?
* view the method invocation details, e.g. method parameter, return object, thrown exception, and etc.
* check the stack trace of specified method invocation. This is useful when you would like to know who is calling your method?
* trace the method invocation to find slow sub-invocations.
* monitoring your system metrics, thread states and cpu usage, gc statistics, and etc.
* command line interactive mode, with auto completed feature enabled.
* telnet and websocket support, which enables both local and remote diagnostics with command line and browsers.
### Quick start
### 快速开始
#### Linux/Unix/Mac
Install Arthas:
安装Arthas:
```
curl -L https://alibaba.github.io/arthas/install.sh | sh
```
Start Arthas:
启动Arthas:
```
./as.sh
@ -46,25 +40,25 @@ Start Arthas:
#### Windows
1. Click [here](http://search.maven.org/classic/#search%7Cga%7C1%7Cg%3A%22com.taobao.arthas%22%20AND%20a%3A%22arthas-packaging%22) to Download the latest binary releases.
Huxing Zhang
6 years ago