yarn 集群部署,遇到的有关问题小结

 

yarn 集群部署,遇到的有关问题小结

优良自学吧提供yarn 集群部署,遇到的有关问题小结,yarn 集群部署,遇到的问题小结 版本信息: hadoop 2.3.0  hive 0.11.0 1. Application Master 无法访问     点击application mater 链接,出现 http 500 错误

 
<iframe src="http://cb.baidu.com/ecom?adn=12&amp;at=103&amp;aurl=&amp;cad=1&amp;ccd=24&amp;cec=UTF-8&amp;cfv=15&amp;ch=0&amp;col=zh-CN&amp;conBW=0&amp;conOP=1&amp;cpa=1&amp;dai=4&amp;dis=0&amp;hn=3&amp;ltr=&amp;ltu=http%3A%2F%2Fwww.ylzx8.cn%2Fzonghe%2Fsystem%2F1013421.html&amp;lunum=6&amp;n=v77y4_cpr&amp;pcs=1635x706&amp;pis=10000x10000&amp;ps=0x0&amp;psr=1920x1080&amp;pss=1635x10768&amp;qn=658e1bd8328454fe&amp;rad=&amp;rs=301&amp;rsi0=450&amp;rsi1=60&amp;rsi5=4&amp;rss0=%23FFFFFF&amp;rss1=%23FFFFFF&amp;rss2=%23FA0303&amp;rss3=&amp;rss4=&amp;rss5=&amp;rss6=%23e10900&amp;rss7=&amp;scale=&amp;skin=&amp;td_id=9223372032564458138&amp;titFF=%E5%AE%8B%E4%BD%93&amp;titFS=14&amp;titTA=left&amp;tn=baiduTlinkInlay&amp;tpr=1418017342632&amp;ts=1&amp;wn=4&amp;xuanting=0&amp;dtm=BAIDU_DUP2_SETJSONADSLOT&amp;dc=2&amp;di=728818" frameborder="0" marginwidth="0" marginheight="0" scrolling="no" align="center,center" width="450" height="60"></iframe>
yarn 集群部署,遇到的问题小结
版本信息: hadoop 2.3.0  hive 0.11.0
 
1. Application Master 无法访问
 
    点击application mater 链接,出现 http 500 错误,java.lang.Connect.exception:
    问题是由于设定web ui时,50030 端口对应的ip地址为0.0.0.0,导致application master 链接无法定位。
 
解决办法:
      yarn-site.xml 文件
    <property>
        <description>The address of the RM web application.</description>
        <name>yarn.resourcemanager.webapp.address</name>
        <value> xxxxxxxxxx:50030</value>
    </property>
    这是2.3.0 的里面的一个bug 1811 ,2.4.0已经修复
 
2. History UI 无法访问 和 container 打不开
     点击 Tracking URL:History无法访问
        问题是 history service 没有启动
      
  解决办法:
     配置:选择( xxxxxxxxxx: 作为history sever
   
    <property>
        <name>yarn.log-aggregation-enable</name>
        <value>true</value>
    </property>
   <property>
        <name>mapreduce.jobhistory.address</name>
        <value> xxxxxxxxxx::10020</value>
    </property>
 
    <property>
    <name>mapreduce.jobhistory.webapp.address</name>
        <value>xxxxxxxxxx:19888</value>
    </property>
 
   sbin/mr-jobhistory-daemon.sh   
start historyserver
相关链接: http://www.iteblog.com/archives/936

3 yarn 平台的优化
 
设置 虚拟cpu的个数
    <property>
        <name>yarn.nodemanager.resource.cpu-vcores</name>
        <value>23</value> 
    </property>
     设置使用的内存
    <property>
        <name>yarn.nodemanager.resource.memory-mb</name>
        <value>61440</value>
        <description>the amount of memory on the NodeManager in GB</description>
    </property>
设置每个任务最大使用的内存
    <property>
        <name>yarn.scheduler.maximum-allocation-mb</name>
        <value>49152</value>
        <source>yarn-default.xml</source>
    </property>

4 运行任务 提示:  Found interface org.apache.hadoop.mapreduce.Counter, but class was expected
修改pom,重新install
    <dependency>
           <groupId>org.apache.hadoop</groupId>
           <artifactId>hadoop-common</artifactId>
           <version>2.3.0</version>
   </dependency>    
 <dependency>
         <groupId>org.apache.hadoop</groupId>
         <artifactId>hadoop-mapreduce-client-core</artifactId>
         <version>2.3.0</version>
   </dependency>
   <dependency>
                <groupId>org.apache.mrunit</groupId>
                <artifactId>mrunit</artifactId>
                <version>1.0.0</version>
                <classifier> hadoop2< /classifier>
                <scope>test</scope>
            </dependency>
jdk 换成1.7



5 运行任务提示shuffle内存溢出 Java heap space
2014-05-14 16:44:22,010 FATAL [IPC Server handler 4 on 44508] org.apache.hadoop.mapred.TaskAttemptListenerImpl: Task: attempt_1400048775904_0006_r_000004_0 - exited : org.apache.hadoop.mapreduce.task.reduce.Shuffle$ShuffleError: error in shuffle in fetcher#3
    at org.apache.hadoop.mapreduce.task.reduce.Shuffle.run(Shuffle.java:134)
    at org.apache.hadoop.mapred.ReduceTask.run(ReduceTask.java:376)
    at org.apache.hadoop.mapred.YarnChild$2.run(YarnChild.java:168)
    at java.security.AccessController.doPrivileged(Native Method)
    at javax.security.auth.Subject.doAs(Subject.java:415)
    at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1548)
    at org.apache.hadoop.mapred.YarnChild.main(YarnChild.java:163)
Caused by: java.lang.OutOfMemoryError: Java heap space
    at org.apache.hadoop.io.BoundedByteArrayOutputStream.<init>(BoundedByteArrayOutputStream.java:56)
    at org.apache.hadoop.io.BoundedByteArrayOutputStream.<init>(BoundedByteArrayOutputStream.java:46)
    at org.apache.hadoop.mapreduce.task.reduce.InMemoryMapOutput.<init>(InMemoryMapOutput.java:63)
    at org.apache.hadoop.mapreduce.task.reduce.MergeManagerImpl.unconditionalReserve(MergeManagerImpl.java:297)
    at org.apache.hadoop.mapreduce.task.reduce.MergeManagerImpl.reserve(MergeManagerImpl.java:287)
    at org.apache.hadoop.mapreduce.task.reduce.Fetcher.copyMapOutput(Fetcher.java:411)
    at org.apache.hadoop.mapreduce.task.reduce.Fetcher.copyFromHost(Fetcher.java:341)
    at org.apache.hadoop.mapreduce.task.reduce.Fetcher.run(Fetcher.java:165)

来源: <http:/xxxxxxxxxx:19888/jobhistory/logs/ST-L09-05-back-tj-yarn15:8034/container_1400048775904_0006_01_000001/job_1400048775904_0006/hadoop/syslog/?start=0>
 
 
解决方法:
调低 mapreduce.reduce.shuffle.memory.limit.percent的值 默认为0.25 现在调成0.10 
 

参考:
http://www.sqlparty.com/yarn%E5%9C%A8shuffle%E9%98%B6%E6%AE%B5%E5%86%85%E5%AD%98%E4%B8%8D%E8%B6%B3%E9%97%AE%E9%A2%98error-in-shuffle-in-fetcher/

6 reduce 任务的log 中间发现:

2014-05-14 17:51:21,835 WARN [Readahead Thread #2] org.apache.hadoop.io.ReadaheadPool: Failed readahead on ifile
EINVAL: Invalid argument
    at org.apache.hadoop.io.nativeio.NativeIO$POSIX.posix_fadvise(Native Method)
    at org.apache.hadoop.io.nativeio.NativeIO$POSIX.posixFadviseIfPossible(NativeIO.java:263)
    at org.apache.hadoop.io.nativeio.NativeIO$POSIX$CacheManipulator.posixFadviseIfPossible(NativeIO.java:142)
    at org.apache.hadoop.io.ReadaheadPool$ReadaheadRequestImpl.run(ReadaheadPool.java:206)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
    at java.lang.Thread.run(Thread.java:745)

来源: <http://xxxxxxxxxx:8042/node/containerlogs/container_1400060792764_0001_01_000726/hadoop/syslog/?start=-4096>
 
ps: 错误没有再现,暂无解决方法
 

7 hive 任务

java.lang.InstantiationException: org.antlr.runtime.CommonToken
Continuing ...
java.lang.RuntimeException: failed to evaluate: <unbound>=Class.new();
 
参考:https://issues.apache.org/jira/browse/HIVE-4222s

8 hive 任务自动把join装换mapjoin时内存溢出,解决方法:关闭自动装换,11前的版本默认值为false,后面的为true;
在任务脚本里面加上:set hive.auto.convert.join=false;
或者在hive-site.xml 配上为false;
出错日志:
SLF4J: Actual binding is of type [org.slf4j.impl.Log4jLoggerFactory]
2014-05-15 02:40:58     Starting to launch local task to process map join;      maximum memory = 1011351552
2014-05-15 02:41:00     Processing rows:        200000  Hashtable size: 199999  Memory usage:   110092544       rate:   0.109
2014-05-15 02:41:01     Processing rows:        300000  Hashtable size: 299999  Memory usage:   229345424       rate:   0.227
2014-05-15 02:41:01     Processing rows:        400000  Hashtable size: 399999  Memory usage:   170296368       rate:   0.168
2014-05-15 02:41:01     Processing rows:        500000  Hashtable size: 499999  Memory usage:   285961568       rate:   0.283
2014-05-15 02:41:02     Processing rows:        600000  Hashtable size: 599999  Memory usage:   408727616       rate:   0.404
2014-05-15 02:41:02     Processing rows:        700000  Hashtable size: 699999  Memory usage:   333867920       rate:   0.33
2014-05-15 02:41:02     Processing rows:        800000  Hashtable size: 799999  Memory usage:   459541208       rate:   0.454
2014-05-15 02:41:03     Processing rows:        900000  Hashtable size: 899999  Memory usage:   391524456       rate:   0.387
2014-05-15 02:41:03     Processing rows:        1000000 Hashtable size: 999999  Memory usage:   514140152       rate:   0.508
2014-05-15 02:41:03     Processing rows:        1029052 Hashtable size: 1029052 Memory usage:   546126888       rate:   0.54
2014-05-15 02:41:03     Dump the hashtable into file: file:/tmp/hadoop/hive_2014-05-15_14-40-53_413_3806680380261480764/-local-10002/HashTable-Stage-4/MapJoin-mapfile01--.hashtable
2014-05-15 02:41:06     Upload 1 File to: file:/tmp/hadoop/hive_2014-05-15_14-40-53_413_3806680380261480764/-local-10002/HashTable-Stage-4/MapJoin-mapfile01--.hashtable File size: 68300588
2014-05-15 02:41:06     End of local task; Time Taken: 8.301 sec.
Execution completed successfully
Mapred Local Task Succeeded . Convert the Join into MapJoin
Mapred Local Task Succeeded . Convert the Join into MapJoin
Launching Job 2 out of 2

log出错日志:
2014-05-15 13:52:54,007 FATAL [main] org.apache.hadoop.mapred.YarnChild: Error running child : java.lang.OutOfMemoryError: Java heap space
    at java.io.ObjectInputStream$HandleTable.grow(ObjectInputStream.java:3465)
    at java.io.ObjectInputStream$HandleTable.assign(ObjectInputStream.java:3271)
    at java.io.ObjectInputStream.readOrdinaryObject(ObjectInputStream.java:1789)
    at java.io.ObjectInputStream.readObject0(ObjectInputStream.java:1350)
    at java.io.ObjectInputStream.readObject(ObjectInputStream.java:370)
    at java.util.HashMap.readObject(HashMap.java:1183)
    at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
    at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:57)
    at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
    at java.lang.reflect.Method.invoke(Method.java:606)
    at java.io.ObjectStreamClass.invokeReadObject(ObjectStreamClass.java:1017)
    at java.io.ObjectInputStream.readSerialData(ObjectInputStream.java:1893)
    at java.io.ObjectInputStream.readOrdinaryObject(ObjectInputStream.java:1798)
    at java.io.ObjectInputStream.readObject0(ObjectInputStream.java:1350)
    at java.io.ObjectInputStream.readObject(ObjectInputStream.java:370)
    at org.apache.hadoop.hive.ql.exec.persistence.HashMapWrapper.initilizePersistentHash(HashMapWrapper.java:128)
    at org.apache.hadoop.hive.ql.exec.MapJoinOperator.loadHashTable(MapJoinOperator.java:194)
    at org.apache.hadoop.hive.ql.exec.MapJoinOperator.cleanUpInputFileChangedOp(MapJoinOperator.java:212)
    at org.apache.hadoop.hive.ql.exec.Operator.cleanUpInputFileChanged(Operator.java:1377)
    at org.apache.hadoop.hive.ql.exec.Operator.cleanUpInputFileChanged(Operator.java:1381)

来源: <http://xxxxxxxxxx:19888/jobhistory/logs/ST-L09-10-back-tj-yarn21:8034/container_1400064445468_0013_01_000002/attempt_1400064445468_0013_m_000000_0/hadoop/syslog/?start=0>
 
 
 

9 hive运行时 提示: failed to evaluate: <unbound>=Class.new(); ,升级到0.13.0
参考https://issues.apache.org/jira/browse/HIVE-4222
https://issues.apache.org/jira/browse/HIVE-3739
 
SLF4J: See http://www.slf4j.org/codes.html#multiple_bindings for an explanation.SLF4J: Actual binding is of type [org.slf4j.impl.Log4jLoggerFactory]OKTime taken: 2.28 secondsjava.lang.InstantiationException: org.antlr.runtime.CommonTokenContinuing ...java.lang.RuntimeException: failed to evaluate: <unbound>=Class.new();Continuing ...java.lang.InstantiationException: org.antlr.runtime.CommonTokenContinuing ...java.lang.RuntimeException: failed to evaluate: <unbound>=Class.new();Continuing ...java.lang.InstantiationException: org.antlr.runtime.CommonTokenContinuing ...java.lang.RuntimeException: failed to evaluate: <unbound>=Class.new();Continuing ...java.lang.InstantiationException: org.antlr.runtime.CommonTokenContinuing ...java.lang.RuntimeException: failed to evaluate: <unbound>=Class.new();Continuing ...java.lang.InstantiationException: org.antlr.runtime.CommonTokenContinuing ...

这个应该升级后能解决,不过不知道为什么我升级12.0 和13.0 ,一运行就报错fileNotfundHIVE_PLANxxxxxxxxx 。ps (参考11)应该是我配置有问题,暂无解决方法。



10 hive 创建表或者数据库的时候   Couldnt obtain a new sequence (unique id) : You have an error in your SQL syntax
解决方法:这个是因为hive元数据库的名字是yarn-hive, sql中中划线是关键词,所以sql错误。把数据库名去掉中划线,问题解决。
错误日志:
FAILED: Error in metadata: MetaException(message:javax.jdo.JDOException: Couldnt obtain a new sequence (unique id) : You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near '-hive.`SEQUENCE_TABLE` WHERE `SEQUENCE_NAME`='org.apache.hadoop.hive.metastore.m' at line 1
        at org.datanucleus.api.jdo.NucleusJDOHelper.getJDOExceptionForNucleusException(NucleusJDOHelper.java:596)
        at org.datanucleus.api.jdo.JDOPersistenceManager.jdoMakePersistent(JDOPersistenceManager.java:732)
        at org.datanucleus.api.jdo.JDOPersistenceManager.makePersistent(JDOPersistenceManager.java:752)
        at org.apache.hadoop.hive.metastore.ObjectStore.createTable(ObjectStore.java:643)
        at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
        at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:57)
        at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
        at java.lang.reflect.Method.invoke(Method.java:606)
        at org.apache.hadoop.hive.metastore.RetryingRawStore.invoke(RetryingRawStore.java:111)
        at com.sun.proxy.$Proxy14.createTable(Unknown Source)
        at org.apache.hadoop.hive.metastore.HiveMetaStore$HMSHandler.create_table_core(HiveMetaStore.java:1070)
        at org.apache.hadoop.hive.metastore.HiveMetaStore$HMSHandler.create_table_with_environment_context(HiveMetaStore.java:1103)
        at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
        at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:57)
        at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
        at java.lang.reflect.Method.invoke(Method.java:606)
        at org.apache.hadoop.hive.metastore.RetryingHMSHandler.invoke(RetryingHMSHandler.java:103)
        at com.sun.proxy.$Proxy15.create_table_with_environment_context(Unknown Source)
        at org.apache.hadoop.hive.metastore.HiveMetaStoreClient.createTable(HiveMetaStoreClient.java:466)
        at org.apache.hadoop.hive.metastore.HiveMetaStoreClient.createTable(HiveMetaStoreClient.java:455)
        at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
        at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:57)
        at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
        at java.lang.reflect.Method.invoke(Method.java:606)
        at org.apache.hadoop.hive.metastore.RetryingMetaStoreClient.invoke(RetryingMetaStoreClient.java:74)
        at com.sun.proxy.$Proxy16.createTable(Unknown Source)
        at org.apache.hadoop.hive.ql.metadata.Hive.createTable(Hive.java:597)
        at org.apache.hadoop.hive.ql.exec.DDLTask.createTable(DDLTask.java:3777)
        at org.apache.hadoop.hive.ql.exec.DDLTask.execute(DDLTask.java:256)
        at org.apache.hadoop.hive.ql.exec.Task.executeTask(Task.java:144)
        at org.apache.hadoop.hive.ql.exec.TaskRunner.runSequential(TaskRunner.java:57)
        at org.apache.hadoop.hive.ql.Driver.launchTask(Driver.java:1362)
        at org.apache.hadoop.hive.ql.Driver.execute(Driver.java:1146)
        at org.apache.hadoop.hive.ql.Driver.run(Driver.java:952)
        at shark.SharkCliDriver.processCmd(SharkCliDriver.scala:338)
        at org.apache.hadoop.hive.cli.CliDriver.processLine(CliDriver.java:413)
        at shark.SharkCliDriver$.main(SharkCliDriver.scala:235)
        at shark.SharkCliDriver.main(SharkCliDriver.scala)
NestedThrowablesStackTrace:
com.mysql.jdbc.exceptions.jdbc4.MySQLSyntaxErrorException: You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near '-hive.`SEQUENCE_TABLE` WHERE `SEQUENCE_NAME`='org.apache.hadoop.hive.metastore.m' at line 1
        at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)
        at sun.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:57)
        at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:45)
        at java.lang.reflect.Constructor.newInstance(Constructor.java:526)
        at com.mysql.jdbc.Util.handleNewInstance(Util.java:406)
        at com.mysql.jdbc.Util.getInstance(Util.java:381)
        at com.mysql.jdbc.SQLError.createSQLException(SQLError.java:1030)
        at com.mysql.jdbc.SQLError.createSQLException(SQLError.java:956)
        at com.mysql.jdbc.MysqlIO.checkErrorPacket(MysqlIO.java:3558)
        at com.mysql.jdbc.MysqlIO.checkErrorPacket(MysqlIO.java:3490)
        at com.mysql.jdbc.MysqlIO.sendCommand(MysqlIO.java:1959)
        at com.mysql.jdbc.MysqlIO.sqlQueryDirect(MysqlIO.java:2109)
        at com.mysql.jdbc.ConnectionImpl.execSQL(ConnectionImpl.java:2648)
        at com.mysql.jdbc.PreparedStatement.executeInternal(PreparedStatement.java:2077)
        at com.mysql.jdbc.PreparedStatement.executeQuery(PreparedStatement.java:2228)
        at org.apache.commons.dbcp.DelegatingPreparedStatement.executeQuery(DelegatingPreparedStatement.java:96)
        at org.apache.commons.dbcp.DelegatingPreparedStatement.executeQuery(DelegatingPreparedStatement.java:96)
        at org.datanucleus.store.rdbms.ParamLoggingPreparedStatement.executeQuery(ParamLoggingPreparedStatement.java:381)
        at org.datanucleus.store.rdbms.SQLController.executeStatementQuery(SQLController.java:504)
        at org.datanucleus.store.rdbms.valuegenerator.SequenceTable.getNextVal(SequenceTable.java:197)
        at org.datanucleus.store.rdbms.valuegenerator.TableGenerator.reserveBlock(TableGenerator.java:190)
        at org.datanucleus.store.valuegenerator.AbstractGenerator.reserveBlock(AbstractGenerator.java:305)
        at org.datanucleus.store.rdbms.valuegenerator.AbstractRDBMSGenerator.obtainGenerationBlock(AbstractRDBMSGenerator.java:170)
        at org.datanucleus.store.valuegenerator.AbstractGenerator.obtainGenerationBlock(AbstractGenerator.java:197)
        at org.datanucleus.store.valuegenerator.AbstractGenerator.next(AbstractGenerator.java:105)
        at org.datanucleus.store.rdbms.RDBMSStoreManager.getStrategyValueForGenerator(RDBMSStoreManager.java:2019)
        at org.datanucleus.store.AbstractStoreManager.getStrategyValue(AbstractStoreManager.java:1385)
        at org.datanucleus.ExecutionContextImpl.newObjectId(ExecutionContextImpl.java:3727)
        at org.datanucleus.state.JDOStateManager.setIdentity(JDOStateManager.java:2574)
        at org.datanucleus.state.JDOStateManager.initialiseForPersistentNew(JDOStateManager.java:526)
        at org.datanucleus.state.ObjectProviderFactoryImpl.newForPersistentNew(ObjectProviderFactoryImpl.java:202)
        at org.datanucleus.ExecutionContextImpl.newObjectProviderForPersistentNew(ExecutionContextImpl.java:1326)
        at org.datanucleus.ExecutionContextImpl.persistObjectInternal(ExecutionContextImpl.java:2123)
        at org.datanucleus.ExecutionContextImpl.persistObjectWork(ExecutionContextImpl.java:1972)
        at org.datanucleus.ExecutionContextImpl.persistObject(ExecutionContextImpl.java:1820)
        at org.datanucleus.ExecutionContextThreadedImpl.persistObject(ExecutionContextThreadedImpl.java:217)
        at org.datanucleus.api.jdo.JDOPersistenceManager.jdoMakePersistent(JDOPersistenceManager.java:727)
        at org.datanucleus.api.jdo.JDOPersistenceManager.makePersistent(JDOPersistenceManager.java:752)
        at org.apache.hadoop.hive.metastore.ObjectStore.createTable(ObjectStore.java:643)


11 安装hive 12 和13 后,运行任务报错提示: FileNotFoundException: HIVE_PLAN
解决方法:可能是hive一个bug,也可能那里配置错了 ,待解决

错误日志

2014-05-16 10:27:07,896 INFO [main] org.apache.hadoop.mapred.MapTask: Processing split: Paths:/user/hive/warehouse/game_predata.db/game_login_log/dt=0000-00-00/000000_0:201326592+60792998,/user/hive/warehouse/game_predata.db/game_login_log/dt=0000-00-00/000001_0_copy_1:201326592+58503492,/user/hive/warehouse/game_predata.db/game_login_log/dt=0000-00-00/000001_0_copy_2:67108864+67108864,/user/hive/warehouse/game_predata.db/game_login_log/dt=0000-00-00/000001_0_copy_2:134217728+67108864,/user/hive/warehouse/game_predata.db/game_login_log/dt=0000-00-00/000002_0_copy_1:67108864+67108864InputFormatClass: org.apache.hadoop.mapred.TextInputFormat

2014-05-16 10:27:07,954 WARN [main] org.apache.hadoop.mapred.YarnChild: Exception running child : java.lang.RuntimeException: java.io.FileNotFoundException: HIVE_PLAN14c8af69-0156-4633-9273-6a812eb91a4c (没有那个文件或目录)
    at org.apache.hadoop.hive.ql.exec.Utilities.getMapRedWork(Utilities.java:230)
    at org.apache.hadoop.hive.ql.io.HiveInputFormat.init(HiveInputFormat.java:255)
    at org.apache.hadoop.hive.ql.io.HiveInputFormat.pushProjectionsAndFilters(HiveInputFormat.java:381)
    at org.apache.hadoop.hive.ql.io.HiveInputFormat.pushProjectionsAndFilters(HiveInputFormat.java:374)
    at org.apache.hadoop.hive.ql.io.CombineHiveInputFormat.getRecordReader(CombineHiveInputFormat.java:540)
    at org.apache.hadoop.mapred.MapTask$TrackedRecordReader.<init>(MapTask.java:168)
    at org.apache.hadoop.mapred.MapTask.runOldMapper(MapTask.java:409)
    at org.apache.hadoop.mapred.MapTask.run(MapTask.java:342)
    at org.apache.hadoop.mapred.YarnChild$2.run(YarnChild.java:168)
    at java.security.AccessController.doPrivileged(Native Method)
    at javax.security.auth.Subject.doAs(Subject.java:415)
    at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1548)
    at org.apache.hadoop.mapred.YarnChild.main(YarnChild.java:163)
Caused by: java.io.FileNotFoundException: HIVE_PLAN14c8af69-0156-4633-9273-6a812eb91a4c (没有那个文件或目录)
    at java.io.FileInputStream.open(Native Method)
    at java.io.FileInputStream.<init>(FileInputStream.java:146)
    at java.io.FileInputStream.<init>(FileInputStream.java:101)
    at org.apache.hadoop.hive.ql.exec.Utilities.getMapRedWork(Utilities.java:221)
    ... 12 more

2014-05-16 10:27:07,957 INFO [main] org.apache.hadoop.mapred.Task: Runnning cleanup for the task

来源: <http://sxxxxxxxxxx:19888/jobhistory/logs/ST-L10-10-back-tj-yarn10:8034/container_1400136017046_0026_01_000030/attempt_1400136017046_0026_m_000000_0/hadoop>
 

12 java.lang.OutOfMemoryError: GC overhead limit exceeded  
分析:这个是JDK6新添的错误类型。是发生在GC占用大量时间为释放很小空间的时候发生的,是一种保护机制。解决方案是,关闭该功能,可以添加JVM的启动参数来限制使用内存: -XX:-UseGCOverheadLimit 
添加位置是:mapred-site.xml 里新增项:mapred.child.java.opts 内容:-XX:-UseGCOverheadLimit
来源: <http://www.cnblogs.com/niocai/archive/2012/07/31/2616252.html>
参考14 
 

13 hive   hive 0.10.0为了执行效率考虑,简单的查询,就是只是select,不带count,sum,group by这样的,都不走map/reduce,直接读取hdfs文件进行filter过滤。这样做的好处就是不新开mr任务,执行效率要提高不少,但是不好的地方就是用户界面不友好,有时候数据量大还是要等很长时间,但是又没有任何返回。

改这个很简单,在hive-site.xml里面有个配置参数叫

hive.fetch.task.conversion

将这个参数设置为more,简单查询就不走map/reduce了,设置为minimal,就任何简单select都会走map/reduce。


来源: <http://slaytanic.blog.51cto.com/2057708/1170431>
 参考14 

14 运行mr 任务的时候提示:

错误日志
Container [pid=30486,containerID=container_1400229396615_0011_01_000012] is running beyond physical memory limits. Current usage: 1.0 GB of 1 GB physical memory used; 1.7 GB of 2.1 GB virtual memory used. Killing container. Dump of the process-tree for container_1400229396615_0011_01_000012 : |- PID PPID PGRPID SESSID CMD_NAME USER_MODE_TIME(MILLIS) SYSTEM_TIME(MILLIS) VMEM_USAGE(BYTES) RSSMEM_USAGE(PAGES) FULL_CMD_LINE |- 30501 30486 30486 30486 (java) 3924 322 1720471552 262096 /opt/jdk1.7.0_55/bin/java -Djava.net.preferIPv4Stack=true -Dhadoop.metrics.log.level=WARN -Xmx1024m -XX:-UseGCOverheadLimit -Djava.io.tmpdir=/home/nodemanager/local/usercache/hadoop/appcache/application_1400229396615_0011/container_1400229396615_0011_01_000012/tmp -Dlog4j.configuration=container-log4j.properties -Dyarn.app.container.log.dir=/home/hadoop/logs/nodemanager/logs/application_1400229396615_0011/container_1400229396615_0011_01_000012 -Dyarn.app.container.log.filesize=0 -Dhadoop.root.logger=INFO,CLA org.apache.hadoop.mapred.YarnChild 30.30.30.39 47925 attempt_1400229396615_0011_m_000000_0 12 |- 30486 12812 30486 30486 (bash) 0 0 108642304 302 /bin/bash -c /opt/jdk1.7.0_55/bin/java -Djava.net.preferIPv4Stack=true -Dhadoop.metrics.log.level=WARN -Xmx1024m -XX:-UseGCOverheadLimit -Djava.io.tmpdir=/home/nodemanager/local/usercache/hadoop/appcache/application_1400229396615_0011/container_1400229396615_0011_01_000012/tmp -Dlog4j.configuration=container-log4j.properties -Dyarn.app.container.log.dir=/home/hadoop/logs/nodemanager/logs/application_1400229396615_0011/container_1400229396615_0011_01_000012 -Dyarn.app.container.log.filesize=0 -Dhadoop.root.logger=INFO,CLA org.apache.hadoop.mapred.YarnChild 30.30.30.39 47925 attempt_1400229396615_0011_m_000000_0 12 1>/home/hadoop/logs/nodemanager/logs/application_1400229396615_0011/container_1400229396615_0011_01_000012/stdout 2>/home/hadoop/logs/nodemanager/logs/application_1400229396615_0011/container_1400229396615_0011_01_000012/stderr Container killed on request. Exit code is 143 Container exited with a non-zero exit code 143


来源: <http://xxxxxxxxxx:50030/proxy/application_1400229396615_0011/mapreduce/attempts/job_1400229396615_0011/m/FAILED>
 
解决方法:
 
下面的参数是关于mapreduce任务运行时的内存设置,如果有的任务需要可单独配置,就统一配置了。如果有container被kill 可以适当调高
mapreduce.map.memory.mb    map任务的最大内存
mapreduce.map.java.opts -Xmx1024M map任务jvm的参数
mapreduce.reduce.memory.mb  reduce任务的最大内存
mapreduce.reduce.java.opts -Xmx2560M reduce任务jvm的参数
mapreduce.task.io.sort.mb 512 Higher memory-limit while sorting data for efficiency.
 
摘自:http://hadoop.apache.org/docs/current/hadoop-project-dist/hadoop-common/ClusterSetup.html#Configuring_the_Hadoop_Daemons_in_Non-Secure_Mode
关闭内存检测进程:
是在搞不清楚 问什么有的任务就物理内存200多MB ,虚拟内存就飙到2.7G了,估计内存检测进程有问题,而且我有的任务是需要大内存的,为了进度,索性关了,一下子解决所有内存问题。
yarn.nodemanager.pmem-check-enabled false
yarn.nodemanager.vmem-check-enabled false
 

15 yarn 的webUI 有关的调整:


1 cluser 页面 application的starttime 和finishtime 都是 UTC格式,改成 +8区时间也就是北京时间。

./share/hadoop/yarn/hadoop-yarn-common-2.3.0.jar 里面的webapps.static. yarn.dt.plugins.js
 
或者源码包里面:/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/resources/webapps/static/yarn.dt.plugins.js

添加代码:
Date.prototype.Format = function (fmt) { //author: meizz 
    var o = {
        "M+": this.getMonth() + 1, //月份 
        "d+": this.getDate(), //日 
        "h+": this.getHours(), //小时 
        "m+": this.getMinutes(), //分 
        "s+": this.getSeconds(), //秒 
        "q+": Math.floor((this.getMonth() + 3) / 3), //季度 
        "S": this.getMilliseconds() //毫秒 
    };
    if (/(y+)/.test(fmt)) fmt = fmt.replace(RegExp.$1, (this.getFullYear() + "").substr(4 - RegExp.$1.length));
    for (var k in o)
    if (new RegExp("(" + k + ")").test(fmt)) fmt = fmt.replace(RegExp.$1, (RegExp.$1.length == 1) ? (o[k]) : (("00" + o[k]).substr(("" + o[k]).length)));
    return fmt;
};

 
同时按下面修改下的代码
function renderHadoopDate(data, type, full) 
{ if (type === 'display' || type === 'filter') { if(data === '0') { return "N/A"; } 
return new Date(parseInt(data)).Format("yyyy-MM-dd hh:mm:ss"); }


 

16  MR1的任务用到DistributedCache 的任务迁移到MR2上出错。原来我里面使用文件名区分不同的缓存文件,MR2里面分发文件以后只保留的文件名如:
application_xxxxxxx/container_14xxxx/part-m-00000
application_xxxxxxx/container_14xxxx/part-m-00001
application_xxxxxxx/container_14xxxx/00000_0



解决方法:每个缓存文件添加符号链接,链接为 父级名字+文件名
DistributedCache.addCacheFile(new URI(path.toString() + "#"+ path.getParent().getName() + "_" + path.getName()),
configuration);


这样就会生成带有文件名的缓存文件

 

 

 

你可能感兴趣的:(yarn)