org.apache.pig.backend.hadoop.executionengine.mapReduceLayer
Interface PigStatsOutputSizeReader

All Known Implementing Classes:

FileBasedOutputSizeReader

@InterfaceAudience.Public
@InterfaceStability.Evolving
public interface PigStatsOutputSizeReader

Interface to implement when you want to customize the way of computing the size of output in PigStats. Since not every storage is file-based (e.g. HBaseStorage), the output size cannot always be computed as the total size of output files.

See Also:

FileBasedOutputSizeReader

Field Summary

static String

OUTPUT_SIZE_READER_KEY

Method Summary

long	getOutputSize(POStore sto, org.apache.hadoop.conf.Configuration conf) Returns the size of output in bytes.
boolean	supports(POStore sto, org.apache.hadoop.conf.Configuration conf) Returns whether the given PSStore is supported by this output size reader or not.

Field Detail

OUTPUT_SIZE_READER_KEY

static final String OUTPUT_SIZE_READER_KEY

See Also:

Constant Field Values

Method Detail

supports

boolean supports(POStore sto,
                 org.apache.hadoop.conf.Configuration conf)

Returns whether the given PSStore is supported by this output size reader or not.

Parameters:

sto - POStore

conf - Configuration

getOutputSize

long getOutputSize(POStore sto,
                   org.apache.hadoop.conf.Configuration conf)
                   throws IOException

Returns the size of output in bytes. If the size of output cannot be computed for any reason, -1 should be returned.

Parameters:

sto - POStore

conf - configuration

Throws:

IOException