org.apache.hadoop.hbase.regionserver

Class TestRowTooBig

java.lang.Object

org.apache.hadoop.hbase.regionserver.TestRowTooBig

public class TestRowTooBig
extends Object

Test case to check HRS throws RowTooBigException when row size exceeds configured limits.

Constructor Summary

Constructors

Constructor and Description
TestRowTooBig()

Method Summary

Modifier and Type	Method and Description
static void	after()
static void	before()
void	testScanAcrossManySmallColumns() Usecase: - create a row with 1M cells, 10 bytes in each - flush & run major compaction - try to Get whole row.
void	testScannersSeekOnFewLargeCells() Usecase: - create a row with 5 large cells (5 Mb each) - flush memstore but don't compact storefiles.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

TestRowTooBig

public TestRowTooBig()

Method Detail

before

public static void before()
                   throws Exception

Throws:

Exception

after

public static void after()
                  throws Exception

Throws:

Exception

testScannersSeekOnFewLargeCells

public void testScannersSeekOnFewLargeCells()
                                     throws IOException

Usecase: - create a row with 5 large cells (5 Mb each) - flush memstore but don't compact storefiles. - try to Get whole row. OOME happened before we actually get to reading results, but during seeking, as each StoreFile gets it's own scanner, and each scanner seeks after the first KV.

Throws:

IOException

testScanAcrossManySmallColumns

public void testScanAcrossManySmallColumns()
                                    throws IOException

Usecase: - create a row with 1M cells, 10 bytes in each - flush & run major compaction - try to Get whole row. OOME happened in StoreScanner.next(..).

Throws:

IOException