上个月,阿里开源了一个名为Arthas的监控工具。恰逢近期自己在写多线程处理业务,由此想到了一个问题。
如果在本机开发调试,IDE可以看到当前的活动线程,例如IntelliJ IDEA,线程是运行还是休眠,在Debugger→Threads可以看到。倘若代码写完提交,到测试环境上构建,要怎样才能看到测试服务器上的线程运行情况呢?
前面客套完了,教程GitHub上面也有,这里主要讲下个人在部署的过程和踩到的坑。仅供参考。
GitHub: https://github.com/alibaba/arthas
Docs: https://alibaba.github.io/arthas/index.html
笔者用的是CentOS7裸机。看过教程,按照步骤执行命令:
curl -L https://alibaba.github.io/arthas/install.sh | sh
下载了个名称为as.sh的脚本,执行脚本,发现:
[root@bogon arthas]# ./as.sh Arthas script version: 3.0.4 illegal ENV, please set $JAVA_HOME to JDK6+
既然是Java监控,没有Java运行环境是不可能跑得起来的(这不是废话吗)。马上安装JDK配置环境变量……这里笔者装的是JDK8u181
JDK: https://www.oracle.com/technetwork/java/javase/downloads/jdk8-downloads-2133151.html
解压到/usr/java,执行vi /etc/profile配置环境变量
export JAVA_HOME=/usr/java/jdk1.8.0_181 export CLASSPATH=.:$JAVA_HOME/lib/dt.jar:$JAVA_HOME/lib/tools.jar:$JAVA_HOME/jre/lib/rt.jar export PATH=$PATH:$JAVA_HOME/bin
:wq保存退出,再执行命令使配置的变量生效
source /etc/profile
回到刚才的目录,启动arthas
[root@bogon ~]# cd /usr/local/arthas/ [root@bogon arthas]# ./as.sh Arthas script version: 3.0.4 Error: no available java process to attach. Usage: ./as.sh [-b [-f SCRIPT_FILE]] [debug] [--use-version VERSION] [--attach-only][@IP:TELNET_PORT:HTTP_PORT] [debug] : start the agent in debug mode : the target Java Process ID [IP] : the target's IP [TELNET_PORT] : the target's PORT for telnet [HTTP_PORT] : the target's PORT for http [-b] : batch mode, which will disable interactive process selection. [-f] : specify the path to batch script file. [--attach-only] : only attach the arthas agent to target jvm. [--use-version] : use the specified arthas version to attach. Example: ./as.sh ./as.sh @[IP] ./as.sh @[IP:PORT] ./as.sh debug ./as.sh -b ./as.sh -b -f /path/to/script ./as.sh --attach-only ./as.sh --use-version 2.0.20161221142407 Here is the list of possible java process(es) to attatch:
结果发现,无可用Java进程。那就启动一个咯(挖鼻屎),继续……
[root@bogon arthas]# ./as.sh Arthas script version: 3.0.4 Found existing java process, please choose one and hit RETURN. * [1]: 1244 test.jar updating version 3.0.4 ... ######################################################################## 100.0% ./as.sh: line 182: unzip: command not found mv: cannot stat ‘/tmp/temp_3.0.4_1265’: No such file or directory update fail, ignore this update. Calculating attach execution time... Attaching to 1244 using version 3.0.4... Error: Unable to access jarfile /root/.arthas/lib/3.0.4/arthas/arthas-core.jar attach to target jvm (1244) failed, check /root/logs/arthas/arthas.log or stderr of target jvm for any exceptions.
然而,程序没有如期运行成功。
……
通过度娘Google各种途径,看了FAQ,作者建议是手动安装+手动拼接命令行启动。但是,有些麻烦,不喜欢(任性)
后面,看到了这个Issue: Error: Unable to access jarfile /root/.arthas/lib/3.0.4/arthas/arthas-core.jar。咦这不就是我共患难的兄弟吗?马上点进去。看到大神ralf0131的回答,顿时茅厕塞顿开:
First, please ensure /root/.arthas/lib/3.0.4/arthas/arthas-core.jar does exist.
cd到/root/.arthas/lib/3.0.4,发现目录下没有jar包。尝不到这瓜有点不甘心,只好手动安装了(真香~)
当前最新的包是3.0.4。下载,解压:
[root@bogon 3.0.4]# wget https://maven.aliyun.com/repository/public/com/taobao/arthas/arthas-packaging/3.0.4/arthas-packaging-3.0.4-bin.zip ...... [root@bogon 3.0.4]# unzip arthas-packaging-3.0.4-bin.zip -bash: unzip: command not found
好吧,我装,我解压……
[root@bogon 3.0.4]# yum -y install unzip [root@bogon 3.0.4]# unzip arthas-packaging-3.0.4-bin.zip
顿时发现,刚才不是说/root/.arthas/lib/3.0.4/arthas路径下木有jar包吗,刚刚我把jar包解压到/root/.arthas/lib/3.0.4,只要再新建个arthas目录,再把4个jar包移进去就OK了。
[root@bogon 3.0.4]# mkdir arthas [root@bogon 3.0.4]# mv *.jar arthas/ [root@bogon arthas]# ./as.sh
大功告……你以为这样就结束了?还有问题呢……
Arthas script version: 3.0.4 Found existing java process, please choose one and hit RETURN. * [1]: 1244 test.jar Calculating attach execution time... Attaching to 1244 using version 3.0.4... real 0m0.620s user 0m0.148s sys 0m0.043s Attach success. Connecting to arthas server... current timestamp is 1539615611 ./as.sh: line 423: type: telnet: not found 'telnet' is required.
还要装telnet-client……
[root@bogon 3.0.4]yum -y install telnet.x86_64
最后再启动,回车,大功告成!
[root@bogon arthas]# ./as.sh Arthas script version: 3.0.4 Found existing java process, please choose one and hit RETURN. * [1]: 1244 test.jar Calculating attach execution time... Attaching to 1244 using version 3.0.4... real 0m0.132s user 0m0.050s sys 0m0.113s Attach success. Connecting to arthas server... current timestamp is 1539615853 Trying 127.0.0.1... Connected to 127.0.0.1. Escape character is '^]'. ,---. ,------. ,--------.,--. ,--. ,---. ,---. / O \ | .--. ''--. .--'| '--' | / O \ ' .-' | .-. || '--'.' | | | .--. || .-. |`. `-. | | | || |\ \ | | | | | || | | |.-' | `--' `--'`--' '--' `--' `--' `--'`--' `--'`-----' wiki: https://alibaba.github.io/arthas version: 3.0.4 pid: 1244 timestamp: 1539615854095 $
接下来就可以各种命令各种姿势各种把弄了……
总结:
-
安装arthas前需要装JDK,配置好JAVA_HOME等环境变量,以及telnet。unzip看情况,如果你解压完再上传到服务器,也行……
-
不知道是bug还是神马情况,按照文档的步骤再一次装在测试的服务器上,程序正常运行,没有报Error: Unable to access jarfile /root/.arthas/lib/3.0.4/arthas/arthas-core.jar的错误,4个jar包都在预期的路径。简单来说,有的机器运行./as.sh是会因少jar包报错的,有的是会正常运行的,具体情况具体分析。后期继续关注arthas的GitHub动态,看项目的最新进展。
就酱。睡觉……