hadoop集群部署报错

gabry 发布于 2014/01/07 21:22
阅读 268
收藏 0

MySQL连接为什么挂死了?别踩坑!>>>

在部署Hadoop过程中发现以下问题,求高手解答。

环境:

master:ip地址192.168.12.44,centos6.4 32位 jdk1.7

slave1:ip地址192.168.12.31,rhel6.4 64位 open-jdk-1.6

hdfs-size.xml等配置文件中都是用IP地址来配置的,不存在IP地址映射的问题。

但是在eclipse调试中报错如下:

14/01/07 20:36:27 INFO mapred.JobClient:  map 94% reduce 1%
14/01/07 20:36:28 INFO mapred.JobClient:  map 96% reduce 1%
14/01/07 20:36:31 INFO mapred.JobClient:  map 97% reduce 1%
14/01/07 20:36:44 INFO mapred.JobClient:  map 100% reduce 1%
14/01/07 20:36:44 INFO mapred.JobClient: Task Id : attempt_201401072030_0001_m_000009_0, Status : FAILED
Too many fetch-failures
14/01/07 20:36:45 WARN mapred.JobClient: Error reading task outputConnection refused: connect
14/01/07 20:36:46 WARN mapred.JobClient: Error reading task outputConnection refused: connect
14/01/07 20:36:53 INFO mapred.JobClient:  map 97% reduce 1%
14/01/07 20:37:17 INFO mapred.JobClient:  map 100% reduce 1%
14/01/07 20:37:38 INFO mapred.JobClient: Task Id : attempt_201401072030_0001_m_000008_0, Status : FAILED
Too many fetch-failures
14/01/07 20:37:39 WARN mapred.JobClient: Error reading task outputConnection refused: connect
14/01/07 20:37:40 WARN mapred.JobClient: Error reading task outputConnection refused: connect
14/01/07 20:37:41 INFO mapred.JobClient:  map 97% reduce 1%

加载中
返回顶部
顶部