Sqoop "import-all-tables" 无法导入所有表

Posted 2023-04-18

【中文标题】Sqoop "import-all-tables" 无法导入所有表

【英文标题】：Sqoop "import-all-tables" unable to import all tables

【发布时间】：2016-10-12 10:31:46

【问题描述】：

这是我用来将数据从 SQL Server 导入 Hive 的 sqoop 命令sqoop-import-all-tables --connect "jdbc:sqlserver://ip.ip.ip.ip\MIGERATIONSERVER;port=1433;username=sa;password=blablaq;database=sqlserverdb" --create-hive-table --hive-import --hive-database hivemtdb 问题是sqlserverdb 有大约 100 个表，但是当我发出此命令时，它只是将 6 或 7 个随机表导入配置单元。这种行为对我来说真的很奇怪。我无法找到我做错的地方。 编辑：1

Warning: /usr/hdp/2.4.3.0-227/accumulo does not exist! Accumulo imports will fail.
Please set $ACCUMULO_HOME to the root of your Accumulo installation.
16/10/13 13:17:38 INFO sqoop.Sqoop: Running Sqoop version: 1.4.6.2.4.3.0-227
16/10/13 13:17:38 INFO tool.BaseSqoopTool: Using Hive-specific delimiters for output. You can override
16/10/13 13:17:38 INFO tool.BaseSqoopTool: delimiters with --fields-terminated-by, etc.
16/10/13 13:17:38 INFO manager.SqlManager: Using default fetchSize of 1000
16/10/13 13:17:38 INFO tool.CodeGenTool: Beginning code generation
16/10/13 13:17:38 INFO manager.SqlManager: Executing SQL statement: SELECT t.* FROM [UserMessage] AS t WHERE 1=0
16/10/13 13:17:38 INFO orm.CompilationManager: HADOOP_MAPRED_HOME is /usr/hdp/2.4.3.0-227/hadoop-mapreduce
Note: /tmp/sqoop-sherry/compile/c809ee201c0aec1edf2ed5a1ef4aed4c/UserMessage.java uses or overrides a deprecated API.
Note: Recompile with -Xlint:deprecation for details.
16/10/13 13:17:39 INFO orm.CompilationManager: Writing jar file: /tmp/sqoop-sherry/compile/c809ee201c0aec1edf2ed5a1ef4aed4c/UserMessage.jar
16/10/13 13:17:39 INFO mapreduce.ImportJobBase: Beginning import of UserMessage
SLF4J: Class path contains multiple SLF4J bindings.
SLF4J: Found binding in [jar:file:/usr/hdp/2.4.3.0-227/hadoop/lib/slf4j-log4j12-1.7.10.jar!/org/slf4j/impl/StaticLoggerBinder.class]
SLF4J: Found binding in [jar:file:/usr/hdp/2.4.3.0-227/zookeeper/lib/slf4j-log4j12-1.6.1.jar!/org/slf4j/impl/StaticLoggerBinder.class]
SLF4J: See http://www.slf4j.org/codes.html#multiple_bindings for an explanation.
SLF4J: Actual binding is of type [org.slf4j.impl.Log4jLoggerFactory]
16/10/13 13:17:40 INFO impl.TimelineClientImpl: Timeline service address: http://machine-02-xx:8188/ws/v1/timeline/
16/10/13 13:17:40 INFO client.RMProxy: Connecting to ResourceManager at machine-02-xx/xxx.xx.xx.xx:8050
16/10/13 13:17:42 INFO db.DBInputFormat: Using read commited transaction isolation
16/10/13 13:17:42 INFO mapreduce.JobSubmitter: number of splits:1
16/10/13 13:17:42 INFO mapreduce.JobSubmitter: Submitting tokens for job: job_1475746531098_0317
16/10/13 13:17:43 INFO impl.YarnClientImpl: Submitted application application_1475746531098_0317
16/10/13 13:17:43 INFO mapreduce.Job: The url to track the job: http://machine-02-xx:8088/proxy/application_1475746531098_0317/
16/10/13 13:17:43 INFO mapreduce.Job: Running job: job_1475746531098_0317
16/10/13 13:17:48 INFO mapreduce.Job: Job job_1475746531098_0317 running in uber mode : false
16/10/13 13:17:48 INFO mapreduce.Job:  map 0% reduce 0%
16/10/13 13:17:52 INFO mapreduce.Job:  map 100% reduce 0%
16/10/13 13:17:52 INFO mapreduce.Job: Job job_1475746531098_0317 completed successfully
16/10/13 13:17:52 INFO mapreduce.Job: Counters: 30
        File System Counters
                FILE: Number of bytes read=0
                FILE: Number of bytes written=156179
                FILE: Number of read operations=0
                FILE: Number of large read operations=0
                FILE: Number of write operations=0
                HDFS: Number of bytes read=87
                HDFS: Number of bytes written=0
                HDFS: Number of read operations=4
                HDFS: Number of large read operations=0
                HDFS: Number of write operations=2
        Job Counters
                Launched map tasks=1
                Other local map tasks=1
                Total time spent by all maps in occupied slots (ms)=3486
                Total time spent by all reduces in occupied slots (ms)=0
                Total time spent by all map tasks (ms)=1743
                Total vcore-seconds taken by all map tasks=1743
                Total megabyte-seconds taken by all map tasks=2677248
        Map-Reduce Framework
                Map input records=0
                Map output records=0
                Input split bytes=87
                Spilled Records=0
                Failed Shuffles=0
                Merged Map outputs=0
                GC time elapsed (ms)=30
                CPU time spent (ms)=980
                Physical memory (bytes) snapshot=233308160
                Virtual memory (bytes) snapshot=3031945216
                Total committed heap usage (bytes)=180879360
        File Input Format Counters
                Bytes Read=0
        File Output Format Counters
                Bytes Written=0
16/10/13 13:17:52 INFO mapreduce.ImportJobBase: Transferred 0 bytes in 12.6069 seconds (0 bytes/sec)
16/10/13 13:17:52 INFO mapreduce.ImportJobBase: Retrieved 0 records.
16/10/13 13:17:52 INFO manager.SqlManager: Executing SQL statement: SELECT t.* FROM [UserMessage] AS t WHERE 1=0
16/10/13 13:17:52 WARN hive.TableDefWriter: Column SendDate had to be cast to a less precise type in Hive
16/10/13 13:17:52 INFO hive.HiveImport: Loading uploaded data into Hive

Logging initialized using configuration in jar:file:/usr/hdp/2.4.3.0-227/hive/lib/hive-common-1.2.1000.2.4.3.0-227.jar!/hive-log4j.properties
OK
Time taken: 1.286 seconds
Loading data to table sqlcmc.usermessage
Table sqlcmc.usermessage stats: [numFiles=1, totalSize=0]
OK
Time taken: 0.881 seconds
Note: /tmp/sqoop-sherry/compile/c809ee201c0aec1edf2ed5a1ef4aed4c/DadChMasConDig.java uses or overrides a deprecated API.
Note: Recompile with -Xlint:deprecation for details.

Logging initialized using configuration in jar:file:/usr/hdp/2.4.3.0-227/hive/lib/hive-common-1.2.1000.2.4.3.0-227.jar!/hive-log4j.properties
OK

【问题讨论】：

将--verbose（检查扩展日志）放入你的命令中，检查是否有任何错误/异常

是的，我也用 --verbose 尝试过，但没有显示任何异常或错误。

试试sqoop list-tables --connect "jdbc:sqlserver://ip.ip.ip.ip\MIGERATIONSERVER;port=1433;username=sa;password=blablaq;database=sqlserverdb"。它显示了所有 100 个表吗？

是的，它显示了所有表格。但我看到它显示了所有表的列表，我意识到它一直只导入前 6 个表（从我现在看到的列表中）

mapreduce 是否有问题，因为它 sqoop 在后端使用 MR，我必须传递特定参数才能将其打开以导入所有表。

【参考方案1】：

首先import-all-tables 将为所有表运行导入表。

如果您没有定义作业中的映射器数量，Sqoop 将默认选择 4 个映射器。所以，它需要表有主键或者你指定--split-by列名。

如果是这种情况，您将看到如下错误：

ERROR tool.ImportAllTablesTool：导入时出错：找不到表测试的主键。请使用 --split-by 指定一个或使用 '-m 1' 执行顺序导入。

因此您可以使用 1 个映射器，这会使您的导入过程变慢。

更好的方法是添加--autoreset-to-one-mapper，它将使用命令中提到的映射器数量导入具有主键的表，并且它将自动为没有主键的表使用1个映射器。

---

来解决你的问题，

表 DadChMasConDig 的 sqoop 导入失败。

不知道为什么没有登录控制台。

在导入此表时可能会出现异常

运行导入作业时遇到 IOException: java.io.IOException: Hive 不支持列 <somecolumn> 的 SQL 类型

例如，varbinary 不受支持。

如果您只在 HDFS 中导入数据，那应该没有问题。你可以试试：

sqoop-import-all-tables --connect "jdbc:sqlserver://ip.ip.ip.ip\MIGERATIONSERVER;port=1433;username=sa;password=blablaq;database=sqlserverdb"

【参考方案2】：

我遇到了同样的问题，以下对我有用。虽然通常 --create-hive-table 和 --hive-overwrite 不会一起使用并且一起没有意义。但是没有其他组合有效，每次只有 10 个表中的 3 个或一小部分表被导入

 sqoop import-all-tables \
       --connect jdbc:mysql://<mysql-url>/my_database \
       --username sql_user \
       --password sql_pwd \
       --hive-import \
       --hive-database test_hive \
       --hive-overwrite \
       --create-hive-table \
       --warehouse-dir /apps/hive/warehouse/test_hive.db \
       -m 1