一个sqoop export案例中踩到的坑

案例分析：

　　需要将hdfs上的数据导出到mysql里的一张表里。

　　虚拟机集群的为：centos1-centos5

问题1：
在centos1上将hdfs上的数据导出到centos1上的mysql里：

sqoop export
　　--connect jdbc:mysql://centos1:3306/test \
　　--username root \
　　--password root \
　　--table order_uid \
　　--export-dir /user/hive/warehouse/test.db/order_uid/ \
　　--fields-terminated-by ','

报错：

Error executing statement: java.sql.SQLException: Access denied for user 'root'@'centos1' (using password: YES)

改成：

sqoop export 
　　--connect jdbc:mysql://localhost:3306/test \
　　--username root \
　　--password root \
　　--table order_uid \
　　--export-dir /user/hive/warehouse/test.db/order_uid/ \
　　--fields-terminated-by ','

报错：

Error: java.io.IOException: com.mysql.jdbc.exceptions.jdbc4.MySQLSyntaxErrorException: Table 'test.order_uid' doesn't exist at
org.apache.sqoop.mapreduce.AsyncSqlRecordWriter.close(AsyncSqlRecordWriter.java:) at 
org.apache.hadoop.mapred.MapTask$NewDirectOutputCollector.close(MapTask.java:) at
org.apache.hadoop.mapred.MapTask.runNewMapper(MapTask.java:) at org.apache.hadoop.mapred.MapTask.run(MapTask.java:) at
org.apache.hadoop.mapred.YarnChild$.run(YarnChild.java:) at java.security.AccessController.doPrivileged(Native Method) at 
javax.security.auth.Subject.doAs(Subject.java:) at
...

问题2：

在centos3上将hdfs上的数据导出到centos1上的mysql里：

sqoop export
　　--connect jdbc:mysql://centos1:3306/test \
　　--username root \
　　--password root \
　　--table order_uid \
　　--export-dir /user/hive/warehouse/test.db/order_uid/ \
　　--fields-terminated-by ','

报错：

// :: ERROR mapreduce.ExportJobBase: Export job failed!
// :: ERROR tool.ExportTool: Error during export:
Export job failed!
at org.apache.sqoop.mapreduce.ExportJobBase.runExport(ExportJobBase.java:)
at org.apache.sqoop.manager.SqlManager.exportTable(SqlManager.java:)
at org.apache.sqoop.tool.ExportTool.exportTable(ExportTool.java:)
at org.apache.sqoop.tool.ExportTool.run(ExportTool.java:)
at org.apache.sqoop.Sqoop.run(Sqoop.java:)
at org.apache.hadoop.util.ToolRunner.run(ToolRunner.java:)
at org.apache.sqoop.Sqoop.runSqoop(Sqoop.java:)
at org.apache.sqoop.Sqoop.runTool(Sqoop.java:)
at org.apache.sqoop.Sqoop.runTool(Sqoop.java:)
at org.apache.sqoop.Sqoop.main(Sqoop.java:)

在网上找到两种解决方案：

1.在网上找到有人说在hdfs的路径写到具体文件，而不是写到目录，改成：

sqoop export
　　--connect jdbc:mysql://centos1:3306/test \
　　--username root \
　　--password root \
　　--table order_uid \
　　--export-dir /user/hive/warehouse/test.db/order_uid/t1.dat \
　　--fields-terminated-by ','

还是报相同错误！

2. 更改mysql里表的编码

将cengos1里mysql的表order_uid字符集编码改成：utf-8，重新执行，centos1的mysql表里导入了部分数据， 仍然报错：

Job failed as tasks failed. failedMaps: failedReduces:
// :: INFO mapreduce.Job: Counters:
File System Counters
FILE: Number of bytes read=
FILE: Number of bytes written=
FILE: Number of read operations=
FILE: Number of large read operations=
FILE: Number of write operations=
HDFS: Number of bytes read=
HDFS: Number of bytes written=
HDFS: Number of read operations=
HDFS: Number of large read operations=
HDFS: Number of write operations=
Job Counters
Failed map tasks=
Killed map tasks=
Launched map tasks=
Data-local map tasks=
Rack-local map tasks=
Total time spent by all maps in occupied slots (ms)=
Total time spent by all reduces in occupied slots (ms)=
Total time spent by all map tasks (ms)=
Total vcore-milliseconds taken by all map tasks=
Total megabyte-milliseconds taken by all map tasks=
Map-Reduce Framework
Map input records=
Map output records=
Input split bytes=
Spilled Records=
Failed Shuffles=
Merged Map outputs=
GC time elapsed (ms)=
CPU time spent (ms)=
Physical memory (bytes) snapshot=
Virtual memory (bytes) snapshot=
Total committed heap usage (bytes)=
File Input Format Counters
Bytes Read=
File Output Format Counters
Bytes Written=
// :: INFO mapreduce.ExportJobBase: Transferred 1.1035 KB in 223.4219 seconds (5.0577 bytes/sec)
// :: INFO mapreduce.ExportJobBase: Exported  records.
// :: ERROR mapreduce.ExportJobBase: Export job failed!
// :: ERROR tool.ExportTool: Error during export:
Export job failed!
at org.apache.sqoop.mapreduce.ExportJobBase.runExport(ExportJobBase.java:)
at org.apache.sqoop.manager.SqlManager.exportTable(SqlManager.java:)
at org.apache.sqoop.tool.ExportTool.exportTable(ExportTool.java:)
at org.apache.sqoop.tool.ExportTool.run(ExportTool.java:)
at org.apache.sqoop.Sqoop.run(Sqoop.java:)
at org.apache.hadoop.util.ToolRunner.run(ToolRunner.java:)
at org.apache.sqoop.Sqoop.runSqoop(Sqoop.java:)
at org.apache.sqoop.Sqoop.runTool(Sqoop.java:)
at org.apache.sqoop.Sqoop.runTool(Sqoop.java:)
at org.apache.sqoop.Sqoop.main(Sqoop.java:)

当指定map为1时：

sqoop export
　　--connect jdbc:mysql://centos1:3306/test \
　　--username root \
　　--password root \
　　--table order_uid \
　　--export-dir /user/hive/warehouse/test.db/order_uid \
　　--fields-terminated-by ',' \
     --m 

运行成功了！！！

sqoop默认情况下的map数量为4，也就是说这种情况下1个map能运行成功，而多个map会失败。于是将map改为2又试了一遍：

sqoop export
　　--connect jdbc:mysql://centos1:3306/test \
　　--username root \
　　--password root \
　　--table order_uid \
　　--export-dir /user/hive/warehouse/test.db/order_uid \
　　--fields-terminated-by ',' \
     --m 2

执行结果为：

// :: INFO mapreduce.Job: Counters:
        File System Counters
                FILE: Number of bytes read=
                FILE: Number of bytes written=
                FILE: Number of read operations=
                FILE: Number of large read operations=
                FILE: Number of write operations=
                HDFS: Number of bytes read=
                HDFS: Number of bytes written=
                HDFS: Number of read operations=
                HDFS: Number of large read operations=
                HDFS: Number of write operations=
        Job Counters
                Failed map tasks=
                Launched map tasks=
                Data-local map tasks=
                Rack-local map tasks=
                Total time spent by all maps in occupied slots (ms)=
                Total time spent by all reduces in occupied slots (ms)=
                Total time spent by all map tasks (ms)=
                Total vcore-milliseconds taken by all map tasks=
                Total megabyte-milliseconds taken by all map tasks=
        Map-Reduce Framework
                Map input records=
                Map output records=
                Input split bytes=
                Spilled Records=
                Failed Shuffles=
                Merged Map outputs=
                GC time elapsed (ms)=
                CPU time spent (ms)=
                Physical memory (bytes) snapshot=
                Virtual memory (bytes) snapshot=
                Total committed heap usage (bytes)=
        File Input Format Counters
                Bytes Read=
        File Output Format Counters
                Bytes Written=
// :: INFO mapreduce.ExportJobBase: Transferred  bytes in 99.4048 seconds (7.0922 bytes/sec)
// :: INFO mapreduce.ExportJobBase: Exported 5 records.
// :: ERROR mapreduce.ExportJobBase: Export job failed!
// :: ERROR tool.ExportTool: Error during export:
Export job failed!
        at org.apache.sqoop.mapreduce.ExportJobBase.runExport(ExportJobBase.java:)
        at org.apache.sqoop.manager.SqlManager.exportTable(SqlManager.java:)
        at org.apache.sqoop.tool.ExportTool.exportTable(ExportTool.java:)
        at org.apache.sqoop.tool.ExportTool.run(ExportTool.java:)
        at org.apache.sqoop.Sqoop.run(Sqoop.java:)
        at org.apache.hadoop.util.ToolRunner.run(ToolRunner.java:)
        at org.apache.sqoop.Sqoop.runSqoop(Sqoop.java:)
        at org.apache.sqoop.Sqoop.runTool(Sqoop.java:)
        at org.apache.sqoop.Sqoop.runTool(Sqoop.java:)
        at org.apache.sqoop.Sqoop.main(Sqoop.java:)

可以看到它成功的导入了5条记录！！！

通过jobhistory窗口可以看到两个map一个成功，一个执行失败

点击task的name：

成功的任务由centos4节点运行的，失败的task由centos1运行，又回到了问题1，就是centos1不能访问centos1的mysql数据！

最终一个朋友告诉我再centos1上单独添加对centos1的远程访问权限：

grant all privileges on *.* to 'root'@'centos1' identified by 'root' with grant option;
flush privileges;

然后重新运行一下，问题1和问题2都被愉快的解决了！！！

当时在centos1上的mysql里执行了：

GRANT ALL PRIVILEGES ON *.* TO 'root'@'%' IDENTIFIED BY 'root' WITH GRANT OPTION; 

flush privileges;

对其他节点添加了远程访问，但没有对自己添加远程访问权限。