Hadoop 1.2.1 - 多节点集群 - Wordcount 程序的减速器阶段挂起?

2022-01-13 00:00:00 hadoop mapreduce word-count java

我的问题在这里可能听起来多余，但之前问题的解决方案都是临时的.我尝试过的很少，但还没有运气.

My question may sound redundant here but the solution to the earlier questions were all ad-hoc. few I have tried but no luck yet.

实际上，我正在研究 hadoop-1.2.1(在 ubuntu 14 上)，最初我有单节点设置然后我运行了 WordCount 程序成功.然后我根据 this 教程添加了一个节点.它成功启动，没有任何错误，但是现在当我运行相同的 WordCount 程序时，它挂在减少阶段.我查看了任务跟踪器日志，它们如下所示:-

Acutally, I am working on hadoop-1.2.1(on ubuntu 14), Initially I had single node set-up and there I ran the WordCount program succesfully. Then I added one more node to it according to this tutorial. It started successfully, without any errors, But now when I am running the same WordCount program it is hanging in reduce phase. I looked at task-tracker logs, they are as given below :-

INFO org.apache.hadoop.mapred.TaskTracker: LaunchTaskAction (registerTask): attempt_201509110037_0001_m_000002_0 task's state:UNASSIGNED INFO org.apache.hadoop.mapred.TaskTracker: Trying to launch : attempt_201509110037_0001_m_000002_0 which needs 1 slots INFO org.apache.hadoop.mapred.TaskTracker: In TaskLauncher, current free slots : 2 and trying to launch attempt_201509110037_0001_m_000002_0 which needs 1 slots INFO org.apache.hadoop.mapred.JobLocalizer: Initializing user hadoopuser on this TT. INFO org.apache.hadoop.mapred.JvmManager: In JvmRunner constructed JVM ID: jvm_201509110037_0001_m_18975496 INFO org.apache.hadoop.mapred.JvmManager: JVM Runner jvm_201509110037_0001_m_18975496 spawned. INFO org.apache.hadoop.mapred.TaskController: Writing commands to /app/hadoop/tmp/mapred/local/ttprivate/taskTracker/hadoopuser/jobcache/job_201509110037_0001/attempt_201509110037_0001_m_000002_0/taskjvm.sh INFO org.apache.hadoop.mapred.TaskTracker: JVM with ID: jvm_201509110037_0001_m_18975496 given task: attempt_201509110037_0001_m_000002_0 INFO org.apache.hadoop.mapred.TaskTracker: attempt_201509110037_0001_m_000002_0 0.0% hdfs://HadoopMaster:54310/input/file02:25+3 INFO org.apache.hadoop.mapred.TaskTracker: Task attempt_201509110037_0001_m_000002_0 is done. INFO org.apache.hadoop.mapred.TaskTracker: reported output size for attempt_201509110037_0001_m_000002_0 was 6 INFO org.apache.hadoop.mapred.TaskTracker: addFreeSlot : current free slots : 2 INFO org.apache.hadoop.mapred.JvmManager: JVM : jvm_201509110037_0001_m_18975496 exited with exit code 0. Number of tasks it ran: 1 INFO org.apache.hadoop.mapred.TaskTracker: LaunchTaskAction (registerTask): attempt_201509110037_0001_r_000000_0 task's state:UNASSIGNED INFO org.apache.hadoop.mapred.TaskTracker: Trying to launch : attempt_201509110037_0001_r_000000_0 which needs 1 slots INFO org.apache.hadoop.mapred.TaskTracker: In TaskLauncher, current free slots : 2 and trying to launch attempt_201509110037_0001_r_000000_0 which needs 1 slots INFO org.apache.hadoop.io.nativeio.NativeIO: Initialized cache for UID to User mapping with a cache timeout of 14400 seconds. INFO org.apache.hadoop.io.nativeio.NativeIO: Got UserName hadoopuser for UID 10 from the native implementation INFO org.apache.hadoop.mapred.JvmManager: In JvmRunner constructed JVM ID: jvm_201509110037_0001_r_18975496 INFO org.apache.hadoop.mapred.JvmManager: JVM Runner jvm_201509110037_0001_r_18975496 spawned. INFO org.apache.hadoop.mapred.TaskController: Writing commands to /app/hadoop/tmp/mapred/local/ttprivate/taskTracker/hadoopuser/jobcache/job_201509110037_0001/attempt_201509110037_0001_r_000000_0/taskjvm.sh INFO org.apache.hadoop.mapred.TaskTracker: JVM with ID: jvm_201509110037_0001_r_18975496 given task: attempt_201509110037_0001_r_000000_0 INFO org.apache.hadoop.mapred.TaskTracker.clienttrace: src: 127.0.1.1:500, dest: 127.0.0.1:55946, bytes: 6, op: MAPRED_SHUFFLE, cliID: attempt_201509110037_0001_m_000002_0, duration: 7129894 INFO org.apache.hadoop.mapred.TaskTracker: attempt_201509110037_0001_r_000000_0 0.11111112% reduce > copy (1 of 3 at 0.00 MB/s) > INFO org.apache.hadoop.mapred.TaskTracker: attempt_201509110037_0001_r_000000_0 0.11111112% reduce > copy (1 of 3 at 0.00 MB/s) > INFO org.apache.hadoop.mapred.TaskTracker: attempt_201509110037_0001_r_000000_0 0.11111112% reduce > copy (1 of 3 at 0.00 MB/s) > INFO org.apache.hadoop.mapred.TaskTracker: attempt_201509110037_0001_r_000000_0 0.11111112% reduce > copy (1 of 3 at 0.00 MB/s) > INFO org.apache.hadoop.mapred.TaskTracker: attempt_201509110037_0001_r_000000_0 0.11111112% reduce > copy (1 of 3 at 0.00 MB/s) > INFO org.apache.hadoop.mapred.TaskTracker: attempt_201509110037_0001_r_000000_0 0.11111112% reduce > copy (1 of 3 at 0.00 MB/s) >

也在我运行程序的控制台上它挂在 -

Also on the console where I am running the program It hangs at -

00:39:24 WARN mapred.JobClient: Use GenericOptionsParser for parsing the arguments. Applications should implement Tool for the same. 00:39:24 INFO util.NativeCodeLoader: Loaded the native-hadoop library 00:39:24 WARN snappy.LoadSnappy: Snappy native library not loaded 00:39:24 INFO mapred.FileInputFormat: Total input paths to process : 2 00:39:24 INFO mapred.JobClient: Running job: job_201509110037_0001 00:39:25 INFO mapred.JobClient: map 0% reduce 0% 00:39:28 INFO mapred.JobClient: map 100% reduce 0% 00:39:35 INFO mapred.JobClient: map 100% reduce 11%

我的配置文件如下:-

//core-site.xml

<configuration> <property> <name>hadoop.tmp.dir</name> <value>/app/hadoop/tmp</value> <description>A base for other temporary directories.</description> </property> <property> <name>fs.default.name</name> <value>hdfs://HadoopMaster:54310</value> <description>The name of the default file system. A URI whose scheme and authority determine the FileSystem implementation. The uri's scheme determines the config property (fs.SCHEME.impl) naming the FileSystem implementation class. The uri's authority is used to determine the host, port, etc. for a filesystem.</description> </property> </configuration>

//hdfs-site.xml

<configuration> <property> <name>dfs.replication</name> <value>1</value> <description>Default block replication. The actual number of replications can be specified when the file is created. The default is used if replication is not specified in create time. </description> </property> </configuration>

//mapred-site.xml

<configuration> <property> <name>mapred.job.tracker</name> <value>HadoopMaster:54311</value> <description>The host and port that the MapReduce job tracker runs at. If "local", then jobs are run in-process as a single map and reduce task. </description> </property> <property> <name>mapred.reduce.slowstart.completed.maps</name> <value>0.80</value> </property> </configuration>

/etc/hosts

127.0.0.1 localhost 127.0.1.1 M-1947 #HADOOP CLUSTER SETUP 172.50.88.54 HadoopMaster 172.50.88.60 HadoopSlave1 # The following lines are desirable for IPv6 capable hosts ::1 ip6-localhost ip6-loopback fe00::0 ip6-localnet ff00::0 ip6-mcastprefix ff02::1 ip6-allnodes ff02::2 ip6-allrouters

/etc/主机名

M-1947

//大师

Hadoop大师

//奴隶

Hadoop大师

HadoopSlave1

HadoopSlave1

我一直在努力解决它，感谢任何帮助.谢谢！

I have been struggling with it for long, any help is appreciated. Thanks !

推荐答案

解决了.. 虽然同一个问题在论坛上有多个问题，但根据我的验证解决方案是主机名解析中的任何节点集群应该是正确的(而且这个问题不取决于集群的大小).

Got it fixed.. although, the same issue has multiple questions on the forums but the verified solution according to me is that hostname resolution for the any node in the cluster should be correct (moreover this issue doesnot depend upon the size of cluster).

实际上是 dns-lookup 的问题，请确保进行以下更改以解决上述问题 -

Actually it is the issue with dns-lookup, ensure one make the below changes to resolve the above issue -

尝试使用'$ hostname'在每台机器上打印主机名

try printing hostname on each machine using '$ hostname'

检查为每台机器打印的主机名是否与各自机器的主/从文件中的条目相同.

check that the hostname printed for each machine is same as the entry made in master/slaves file for respective machine.

示例:-

在/etc/hosts文件中(比如说在hadoop集群的Master机器上)

in /etc/hosts file (let's say on Master machine of hadoop cluster)

127.0.0.1 本地主机

127.0.0.1 localhost

127.0.1.1 约翰机器

127.0.1.1 john-machine

#Hadoop 集群

172.50.88.21 HadoopMaster

172.50.88.21 HadoopMaster

172.50.88.22 HadoopSlave1

172.50.88.22 HadoopSlave1

172.50.88.23 HadoopSlave2

172.50.88.23 HadoopSlave2

那么它是 -> /etc/hostname 文件(在主机上)应该包含以下条目(用于解决上述问题)

then it's -> /etc/hostname file (on master machine) should contain the following entry (for the above issue to be resolved)

Hadoop大师

类似地验证每个从节点的/etc/hostname 文件.

similarly verify the /etc/hostname files of the each slave node.

相关文章