使用Mongo dump 将数据导入到hive

概述：使用dump 方式将mongo数据导出，上传到hdfs,然后在hive中建立外部表。

1.     使用mongodump 将集合导出

mongodump --host=localhost:27017  --db=mydb --collection=users  --out=/tmp/root/mongodump0712

[root@slave2 root]# mongodump --host=localhost:27017  --db=mydb --collection=users  --out=/tmp/root/mongodump0712 

2018-07-12T10:07:27.894+0800    writing mydb.users to 

2018-07-12T10:07:27.896+0800    done dumping mydb.users (2 documents)

[root@slave2 root]# cd /tmp/root

[root@slave2 root]# ls

3604abd2-a359-4c53-a7b4-e4ea84185801  3604abd2-a359-4c53-a7b4-e4ea841858017799130181720133073.pipeout  dump  hive.log  hive.log.2018-07-11  mongodump0712

[root@slave2 root]# ll

total 624

drwx------. 2 root root      6 Jul 12 09:34 3604abd2-a359-4c53-a7b4-e4ea84185801

-rw-r--r--. 1 root root      0 Jul 12 09:34 3604abd2-a359-4c53-a7b4-e4ea841858017799130181720133073.pipeout

drwxr-xr-x. 5 root root     44 Jul 12 10:04 dump

-rw-r--r--. 1 root root  88700 Jul 12 09:39 hive.log

-rw-r--r--. 1 root root 547126 Jul 11 21:07 hive.log.2018-07-11

drwxr-xr-x. 3 root root     18 Jul 12 10:07 mongodump0712

[root@slave2 root]# cd mongodump0712/

[root@slave2 mongodump0712]# ls

mydb

[root@slave2 mongodump0712]# cd mydb

[root@slave2 mydb]# ls

users.bson  users.metadata.json    

2.     将dump文件上传到hdfs

hdfs dfs -mkdir /user/hive/warehouse/mongo

hdfs dfs -put /tmp/root/mongodump0712/mydb/users.bson /user/hive/warehouse/mongo/

[root@slave2 mydb]# hdfs dfs -mkdir /user/hive/warehouse/mongo 

[root@slave2 mydb]# hdfs dfs -put /tmp/root/mongodump0712/mydb/users.bson /user/hive/warehouse/mongo/  

3.     创建表并测试

hive> create EXTERNAL table muser

    > (

    >   id string,

    >   userid string,

    >   age bigint,

    >   status string

    > )

    > row format serde 'com.mongodb.hadoop.hive.BSONSerDe'

    > WITH SERDEPROPERTIES('mongo.columns.mapping'='{"id":"_id","userid":"user_id","age":"age","status":"status"}')

    > stored as inputformat 'com.mongodb.hadoop.mapred.BSONFileInputFormat'

    > outputformat 'com.mongodb.hadoop.hive.output.HiveBSONFileOutputFormat'

    > location '/user/hive/warehouse/muser';

FAILED: Execution Error, return code 1 from org.apache.hadoop.hive.ql.exec.DDLTask. MetaException(message:hdfs://ns1/user/hive/warehouse/muser is not a directory or unable to create one)

hive> create EXTERNAL table muser

    > (

    >   id string,

    >   userid string,

    >   age bigint,

    >   status string

    > )

    > row format serde 'com.mongodb.hadoop.hive.BSONSerDe'

    > WITH SERDEPROPERTIES('mongo.columns.mapping'='{"id":"_id","userid":"user_id","age":"age","status":"status"}')

    > stored as inputformat 'com.mongodb.hadoop.mapred.BSONFileInputFormat'

    > outputformat 'com.mongodb.hadoop.hive.output.HiveBSONFileOutputFormat'

    > location '/user/hive/warehouse/mongo';

OK

Time taken: 0.123 seconds

hive> select * from muser;

OK

5b456e33a93daf7ae53e6419        abc123  58      D

5b45705ca93daf7ae53e8b2a        bcd001  45      C

Time taken: 0.181 seconds, Fetched: 2 row(s)