默认情况下,我们使用的都是 jre 版本的 openjdk,当容器启动卡住不动的时候,看不出来任何问题。

此时如果能 dump 就能知道线程在干啥,也能找到一些大概的问题。

此时 jre 版本的镜像就不够用了。

切换 jre 为 jdk 版本

只切换为 jdk 还不够,还会遇到 Unable to get pid of LinuxThreads manager thread 的错误。

创建可以 dump 用的基础镜像

参考前面文章,创建如下镜像:

FROM openjdk:8u191-jdk-alpine3.9

RUN apk add --no-cache tini

ENTRYPOINT ["tini"]

修改项目使用的镜像和启动方式

假设上面创建的镜像名为 openjdk:8u191-jdk-alpine3.9-tini

FROM openjdk:8u191-jdk-alpine3.9-tini

COPY app.jar /opt/dubbo-app/app.jar

WORKDIR /opt/dubbo-app

EXPOSE 20880

ENTRYPOINT ["/sbin/tini", "--", "java", "-jar", "app.jar"]

启动镜像后进入容器

jps 查看 pid

jstack -l pid 查看线程信息

关于此次 BUG

经过查看堆栈和代码,发现是 Dubbo 连接 zookeeper 时,用了 CountDownLatch ,由于通过环境变量配置的 ZOOKEEPER 地址中,环境变量名竟然配错了,导致 zookeeper 一直连接不上,因此锁死了主线程。

实际上这里没有添加 timeout 也是 Dubbo 2.7.1 的一大 BUG。

dubbo 2.7.1 有很多严重 BUG,而且修复和发布的周期特别的长,一定要慎用。

主线程堆栈信息:

"main" #1 prio=5 os_prio=0 tid=0x00005592eb0f1000 nid=0x9 waiting on condition [0x00007fda15afd000]

java.lang.Thread.State: WAITING (parking)

at sun.misc.Unsafe.park(Native Method)

- parking to wait for <0x00000000f885dac0> (a java.util.concurrent.CountDownLatch$Sync)

at java.util.concurrent.locks.LockSupport.park(LockSupport.java:175)

at java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:836)

at java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireSharedInterruptibly(AbstractQueuedSynchronizer.java:997)

at java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireSharedInterruptibly(AbstractQueuedSynchronizer.java:1304)

at java.util.concurrent.CountDownLatch.await(CountDownLatch.java:231)

at org.apache.dubbo.configcenter.support.zookeeper.ZookeeperDynamicConfiguration.(ZookeeperDynamicConfiguration.java:64)

at org.apache.dubbo.configcenter.support.zookeeper.ZookeeperDynamicConfigurationFactory.createDynamicConfiguration(ZookeeperDynamicConfigurationFactory.java:38)

at org.apache.dubbo.configcenter.AbstractDynamicConfigurationFactory.getDynamicConfiguration(AbstractDynamicConfigurationFactory.java:33)

- locked <0x00000000f885db68> (a org.apache.dubbo.configcenter.support.zookeeper.ZookeeperDynamicConfigurationFactory)

at org.apache.dubbo.config.AbstractInterfaceConfig.getDynamicConfiguration(AbstractInterfaceConfig.java:275)

at org.apache.dubbo.config.AbstractInterfaceConfig.prepareEnvironment(AbstractInterfaceConfig.java:250)

at org.apache.dubbo.config.AbstractInterfaceConfig.startConfigCenter(AbstractInterfaceConfig.java:240)

at org.apache.dubbo.config.AbstractInterfaceConfig.lambda$null$7(AbstractInterfaceConfig.java:584)

at org.apache.dubbo.config.AbstractInterfaceConfig$$Lambda$218/1961945640.get(Unknown Source)

at java.util.Optional.orElseGet(Optional.java:267)

对应代码截图如下:

docker无法启动seata docker无法启动 jdk_apache