记一次排查tomcat耗费CPU过高的经历

有一个新项目，在测试环境部署后，发现tomcat进程耗费的CPU非常高，排查过程如下：

日志搜集

先通过top，查找耗费CPU最高的线程

top -Hp pid

将线程ID转为16进制

printf "%x\n" threadid

搜集JVM的栈日志

jstack pid > ~/jvm.stack.log

通过线程的16进制标识，定位JVM栈信息

grep -rn threadTag ~/jvm.stack.log -A 100

原因分析

通过上面的操作，定位到具体的线程栈日志，如下：

"pool-8-thread-1" #253 prio=5 os_prio=0 tid=0x00007f538096f000 nid=0x7fdb runnable [0x00007f5363af8000]
    java.lang.Thread.State: RUNNABLE
     at sun.nio.ch.IOUtil.drain(Native Method)
     at sun.nio.ch.EPollSelectorImpl.doSelect(EPollSelectorImpl.java:104)
     - locked <0x00000000f36f0960> (a java.lang.Object)
     at sun.nio.ch.SelectorImpl.lockAndDoSelect(SelectorImpl.java:86)
     - locked <0x00000000f36f0750> (a sun.nio.ch.Util$3)
     - locked <0x00000000f36f0740> (a java.util.Collections$UnmodifiableSet)
     - locked <0x00000000f36f0558> (a sun.nio.ch.EPollSelectorImpl)
     at sun.nio.ch.SelectorImpl.select(SelectorImpl.java:97)
     at org.apache.kafka.common.network.Selector.select(Selector.java:470)
     at org.apache.kafka.common.network.Selector.poll(Selector.java:286)
     at org.apache.kafka.clients.NetworkClient.poll(NetworkClient.java:260)
     at org.apache.kafka.clients.consumer.internals.ConsumerNetworkClient.poll(ConsumerNetworkClient.java:232)
     - locked <0x00000000f37037e8> (a org.apache.kafka.clients.consumer.internals.ConsumerNetworkClient)
     at org.apache.kafka.clients.consumer.internals.ConsumerNetworkClient.poll(ConsumerNetworkClient.java:209)
     at org.apache.kafka.clients.consumer.internals.ConsumerNetworkClient.awaitMetadataUpdate(ConsumerNetworkClient.java:148)
     at org.apache.kafka.clients.consumer.internals.ConsumerNetworkClient.awaitMetadataUpdate(ConsumerNetworkClient.java:136)
     at org.apache.kafka.clients.consumer.internals.AbstractCoordinator.ensureCoordinatorReady(AbstractCoordinator.java:197)
     - locked <0x00000000f3703820> (a org.apache.kafka.clients.consumer.internals.ConsumerCoordinator)
     at org.apache.kafka.clients.consumer.internals.ConsumerCoordinator.poll(ConsumerCoordinator.java:248)
     at org.apache.kafka.clients.consumer.KafkaConsumer.pollOnce(KafkaConsumer.java:1013)
     // 可以看到是kafka的poll出现问题了，触发locked
     at org.apache.kafka.clients.consumer.KafkaConsumer.poll(KafkaConsumer.java:979)
     at com.myhexin.kiv.service.impl.SyncTemplateServiceImpl.receiver(SyncTemplateServiceImpl.java:65)
     at com.myhexin.kiv.job.SchedulerAllJob.registerJobs(SchedulerAllJob.java:105)
     at com.myhexin.kiv.job.SchedulerAllJob.scheduleJobs(SchedulerAllJob.java:68)
     at com.myhexin.kiv.job.SchedulerListener.schedule(SchedulerListener.java:37)
     at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
     at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
     at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
     at java.lang.reflect.Method.invoke(Method.java:498)
     at org.springframework.scheduling.support.ScheduledMethodRunnable.run(ScheduledMethodRunnable.java:65)
     at org.springframework.scheduling.support.DelegatingErrorHandlingRunnable.run(DelegatingErrorHandlingRunnable.java:54)
     at org.springframework.scheduling.concurrent.ReschedulingRunnable.run(ReschedulingRunnable.java:81)
     at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
     at java.util.concurrent.FutureTask.run(FutureTask.java:266)
     at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$201(ScheduledThreadPoolExecutor.java:180)
     at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:293)
     at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
     at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
     at java.lang.Thread.run(Thread.java:745)

从上面的日志可以看出，是kafka的poll出现问题了，触发locked。然后经排查，发现是我们将测试环境的kafka连接参数，误配置成了正式环境的参数；而测试环境和正式环境的网络是隔离的，无法连通，因此出现该问题。