org.apache.hadoop.hive.ql.io
Interface AcidOutputFormat<V>

Type Parameters:

V - the row type of the file

All Superinterfaces:

HiveOutputFormat<org.apache.hadoop.io.NullWritable,V>, org.apache.hadoop.mapred.OutputFormat<org.apache.hadoop.io.NullWritable,V>

All Known Implementing Classes:

OrcOutputFormat

public interface AcidOutputFormat<V>
extends HiveOutputFormat<org.apache.hadoop.io.NullWritable,V>

An extension for OutputFormats that want to implement ACID transactions.

Nested Class Summary

static class

AcidOutputFormat.Options Options to control how the files are written

Method Summary

FileSinkOperator.RecordWriter	getRawRecordWriter(org.apache.hadoop.fs.Path path, AcidOutputFormat.Options options) Create a raw writer for ACID events.
RecordUpdater	getRecordUpdater(org.apache.hadoop.fs.Path path, AcidOutputFormat.Options options) Create a RecordUpdater for inserting, updating, or deleting records.

Methods inherited from interface org.apache.hadoop.hive.ql.io.HiveOutputFormat

getHiveRecordWriter

Methods inherited from interface org.apache.hadoop.mapred.OutputFormat

checkOutputSpecs, getRecordWriter

Method Detail

getRecordUpdater

RecordUpdater getRecordUpdater(org.apache.hadoop.fs.Path path,
                               AcidOutputFormat.Options options)
                               throws IOException

Create a RecordUpdater for inserting, updating, or deleting records.

Parameters:

path - the partition directory name

options - the options for the writer

Returns:

the RecordUpdater for the output file

Throws:

IOException

getRawRecordWriter

FileSinkOperator.RecordWriter getRawRecordWriter(org.apache.hadoop.fs.Path path,
                                                 AcidOutputFormat.Options options)
                                                 throws IOException

Create a raw writer for ACID events. This is only intended for the compactor.

Parameters:

path - the root directory

options - options for writing the file

Returns:

a record writer

Throws:

IOException