|
||||||||||
| PREV PACKAGE NEXT PACKAGE | FRAMES NO FRAMES | |||||||||
See:
Description
| Interface Summary | |
|---|---|
| DBWritable | Objects that are read from/written to a database should implement
DBWritable. |
| Class Summary | |
|---|---|
| DBConfiguration | A container for configuration property names for jobs with DB input/output. |
| DBInputFormat<T extends DBWritable> | A InputFormat that reads input data from an SQL table. |
| DBInputFormat.DBInputSplit | A InputSplit that spans a set of rows |
| DBInputFormat.NullDBWritable | A Class that does nothing, implementing DBWritable |
| DBOutputFormat<K extends DBWritable,V> | A OutputFormat that sends the reduce output to a SQL table. |
This package contains a library to read records from a database as an input to a mapreduce job, and write the output records to the database.
The Database to access can be configured using the static methods in the DBConfiguration class. Jobs reading input from a database should use DBInputFormat#setInput() to set the configuration. And jobs writing its output to the database should use DBOutputFormat#setOutput().
Tuples from/to the database are converted to/from Java objects using DBWritable methods. Typically, for each table in the db, a class extending DBWritable is defined, which holds the fields of the tuple. The fields of a record are read from the database using DBWritable#readFields(ResultSet), and written to the database using DBWritable#write(PreparedStatament statement).
An example program using both DBInputFormat and DBOutputFormat can be found at src/examples/org/apache/hadoop/examples/DBCountPageview.java.
|
||||||||||
| PREV PACKAGE NEXT PACKAGE | FRAMES NO FRAMES | |||||||||