使用Java-Hadoop连接器更新MongoDB中的现有集合

Question

Is it possible to update existing MongoDB collection with new data. I am using hadoop job to read write data to Mongo. Required scenario is :- Say first collection in Mongo is

{ 
  "_id" : 1,
  "value" : "aaa"
  "value2" : null
}

after reading data from Mongo and processing data, MongoDB should contain

{ 
  "_id" : 1,
  "value" : "aaa"
  "value2" : "bbb"
}

If possible, please provide some dummy code.

Answer 1

BasicBSONObject query=new BasicBSONObject();
query.append("fieldname", value);
BasicBSONObject update=new BasicBSONObject();

update.append("$set", new BasicBSONObject().append("newfield",value));
MongoUpdateWritable muw=new MongoUpdateWritable(query,update,false,true);
contex.write(key, muw);

query : is used for providing condition(matching condition).

update : is used for adding new field and value in existing collection.

MongoUpdateWritable : 3rd parameter is upsert value(same as mongodb)

4th parameter is multiple update in many documents in a collection.

Set in Driver class job.setOutputValueClass(MongoUpdateWritable.class);

Answer 2

我通过扩展org.apache.hadoop.mapreduce.RecordWriter并覆盖此类的write方法来完成此操作。

Answer 3

The Mongo-Hadoop Connector doesn't currently suppor this feature. You can open a feature request in the MongoDB Jira if you like.

Answer 4

我已经完成了分层工作 ，如果您正在使用spark，则可以检查一下！