java線程阻塞問題排查方法

  我開發的worker，每隔幾個月線上都會阻塞一次，一直都沒查出問題。今天終於了了這個心結。把解決過程總結下和大家分享。

    首先用jstack命令打出這個進程的全部線程堆棧。拿到線程dump文件之后，搜索自己的worker名字。

 

1. "DefaultQuartzScheduler_Worker-10" prio=10 tid=0x00007f55cd54d800 nid=0x3e2e waiting for monitor entry [0x00007f51ab8f7000]
2. java.lang.Thread.State: BLOCKED ( on object monitor)
3. at com.jd.chat.worker.service.impl.NewPopAccountSyncServiceImpl.addAccounts(NewPopAccountSyncServiceImpl.java: 86)
4. - waiting to lock <0x0000000782359268> (a com.jd.chat.worker.service.impl.NewPopAccountSyncServiceImpl)
5. at com.jd.chat.worker.service. timer.AccountIncSyncTimer.run(AccountIncSyncTimer.java:114)
6. at com.jd.chat.worker.service. timer.AbstractTimer.start(AbstractTimer.java:44)
7. at com.jd.chat.worker.service. timer.AbstractTimer.doJob(AbstractTimer.java:49)
8. at com.jd.chat.worker.web.context.StartAppListener$TimerJob. execute(StartAppListener.java:188)
9. at org.quartz.core.JobRunShell.run(JobRunShell.java: 202)
10. at org.quartz.simpl.SimpleThreadPool$WorkerThread.run(SimpleThreadPool.java: 573)
11. - locked < 0x0000000783641c68> (a java.lang.Object)

    很快便找到了線程在哪一行被阻塞。但是就憑這么點信息，並不能查出問題的真正原因，這里推薦一個工具，叫tda.bat。同事給我的，網上應該有下載。把這個dump文件導入到tda中。找到阻塞的線程。阻塞的線程是紅色的。

   之所以說這個軟件好，是因為當你找到blocked的線程后，界面的下方，會打出阻塞的更詳細的線程堆棧。截取這個堆棧的部分信息。

 

1. at org.mariadb.jdbc.MySQLPreparedStatement.execute(MySQLPreparedStatement.java: 141)
2. at org.apache.commons.dbcp.DelegatingPreparedStatement.execute(DelegatingPreparedStatement.java: 172)
3. at org.apache.commons.dbcp.DelegatingPreparedStatement.execute(DelegatingPreparedStatement.java: 172)
4. at com.ibatis.sqlmap.engine.execution.SqlExecutor.executeUpdate(SqlExecutor.java: 80)
5. at com.ibatis.sqlmap.engine.mapping.statement.MappedStatement.sqlExecuteUpdate(MappedStatement.java: 216)
6. at com.ibatis.sqlmap.engine.mapping.statement.MappedStatement.executeUpdate(MappedStatement.java: 94)
7. at com.ibatis.sqlmap.engine.impl.SqlMapExecutorDelegate.update(SqlMapExecutorDelegate.java: 457)
8. at com.ibatis.sqlmap.engine.impl.SqlMapSessionImpl.update(SqlMapSessionImpl.java: 90)
9. at org.springframework.orm.ibatis.SqlMapClientTemplate$ 9.doInSqlMapClient(SqlMapClientTemplate.java:380)
10. at org.springframework.orm.ibatis.SqlMapClientTemplate$ 9.doInSqlMapClient(SqlMapClientTemplate.java:1)
11. at org.springframework.orm.ibatis.SqlMapClientTemplate.execute(SqlMapClientTemplate.java: 200)
12. at org.springframework.orm.ibatis.SqlMapClientTemplate.update(SqlMapClientTemplate.java: 378)
13. at com.jd.im.data.dataresource.ImSqlMapClientTemplate.retriedWithoutAnyInterventionUpdate(ImSqlMapClientTemplate.java: 169)
14. at com.jd.im.data.dataresource.ImSqlMapClientTemplate.update(ImSqlMapClientTemplate.java: 137)
15. at com.jd.chat.dao.impl.WriteDaoImpl.update(WriteDaoImpl.java: 21)
16. at com.jd.chat.zone.service.impl.GroupServiceImpl.updateRoute(GroupServiceImpl.java: 766)
17. at com.jd.chat.worker.service.impl.NewPopAccountSyncServiceImpl.addAccounts(NewPopAccountSyncServiceImpl.java: 267)
18. - locked < 0x0000000782359268> (a com.jd.chat.worker.service.impl.NewPopAccountSyncServiceImpl)

 

    這個才是真正有用的堆棧！它告訴了我程序是在執行SQL的時候，SQL發生死鎖，於是線程被阻塞。它還提供了更有用的信息，那就是到底是哪個SQL導致的死鎖。堆棧的倒數第三行指示了導致死鎖的SQL。

    但是一定要用這個工具才能找到具體的原因嗎？答案當然是NO！

    告訴大家怎么不通過工具找到阻塞的真正原因！

    剛剛通過“BLOCKED”關鍵字搜到了線程堆棧，找到它的線程名“DefaultQuartzScheduler_Worker-10”。OK，然后，把最后的10改成1，也就是“DefaultQuartzScheduler_Worker-1”，然后再拿這個關鍵字搜索整個進程堆棧。

 

1. "DefaultQuartzScheduler_Worker-1" prio=10 tid=0x00007f55cd2aa000 nid=0x3e25 runnable [0x00007f51b02c0000]
2. java.lang.Thread.State: RUNNABLE
3. at java.net.SocketInputStream.socketRead0(Native Method)
4. at java.net.SocketInputStream.read(SocketInputStream.java: 129)
5. at java.io.BufferedInputStream.fill(BufferedInputStream.java: 218)
6. at java.io.BufferedInputStream.read1(BufferedInputStream.java: 258)
7. at java.io.BufferedInputStream.read(BufferedInputStream.java: 317)
8. - locked < 0x0000000791370d50> (a java.io.BufferedInputStream)
9. at org.mariadb.jdbc.internal.common.packet.buffer.ReadUtil.readFully(ReadUtil.java: 82)
10. at org.mariadb.jdbc.internal.common.packet.buffer.ReadUtil.readFully(ReadUtil.java: 92)
11. at org.mariadb.jdbc.internal.common.packet.RawPacket.nextPacket(RawPacket.java: 77)
12. at org.mariadb.jdbc.internal.common.packet.SyncPacketFetcher.getRawPacket(SyncPacketFetcher.java: 67)
13. at org.mariadb.jdbc.internal.mysql.MySQLProtocol.getResult(MySQLProtocol.java: 891)
14. at org.mariadb.jdbc.internal.mysql.MySQLProtocol.executeQuery(MySQLProtocol.java: 982)
15. at org.mariadb.jdbc.MySQLStatement.execute(MySQLStatement.java: 280)
16. - locked < 0x0000000791370678> (a org.mariadb.jdbc.internal.mysql.MySQLProtocol)
17. at org.mariadb.jdbc.MySQLPreparedStatement.execute(MySQLPreparedStatement.java: 141)

 

               貼出這個進程堆棧的一部分。這個進程堆棧其實也就是剛剛tda軟件界面下方展示的導致線程阻塞的真正的堆棧！這個線程是runnable狀態的，可惜mysql是鎖死的。也就是說阻塞在了mysql里。

    感覺這是一個由張三的命案牽出李四的命案的故事。