001/* 002 * Licensed to the Apache Software Foundation (ASF) under one 003 * or more contributor license agreements. See the NOTICE file 004 * distributed with this work for additional information 005 * regarding copyright ownership. The ASF licenses this file 006 * to you under the Apache License, Version 2.0 (the 007 * "License"); you may not use this file except in compliance 008 * with the License. You may obtain a copy of the License at 009 * 010 * http://www.apache.org/licenses/LICENSE-2.0 011 * 012 * Unless required by applicable law or agreed to in writing, software 013 * distributed under the License is distributed on an "AS IS" BASIS, 014 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 015 * See the License for the specific language governing permissions and 016 * limitations under the License. 017 */ 018package org.apache.hadoop.hbase.regionserver; 019 020import static org.junit.Assert.assertEquals; 021import static org.junit.Assert.assertTrue; 022 023import java.io.IOException; 024import java.util.ArrayList; 025import java.util.Arrays; 026import java.util.Collection; 027import java.util.Collections; 028import java.util.Comparator; 029import java.util.List; 030import java.util.NavigableSet; 031import java.util.TreeSet; 032import org.apache.hadoop.conf.Configuration; 033import org.apache.hadoop.fs.FileSystem; 034import org.apache.hadoop.fs.Path; 035import org.apache.hadoop.hbase.Cell; 036import org.apache.hadoop.hbase.CellUtil; 037import org.apache.hadoop.hbase.HBaseClassTestRule; 038import org.apache.hadoop.hbase.HBaseTestingUtility; 039import org.apache.hadoop.hbase.HColumnDescriptor; 040import org.apache.hadoop.hbase.KeyValue; 041import org.apache.hadoop.hbase.KeyValueTestUtil; 042import org.apache.hadoop.hbase.client.Put; 043import org.apache.hadoop.hbase.client.Scan; 044import org.apache.hadoop.hbase.io.compress.Compression; 045import org.apache.hadoop.hbase.io.hfile.HFilePrettyPrinter; 046import org.apache.hadoop.hbase.testclassification.RegionServerTests; 047import org.apache.hadoop.hbase.testclassification.SmallTests; 048import org.apache.hadoop.hbase.util.BloomFilterUtil; 049import org.apache.hadoop.hbase.util.Bytes; 050import org.junit.Before; 051import org.junit.ClassRule; 052import org.junit.Test; 053import org.junit.experimental.categories.Category; 054import org.junit.runner.RunWith; 055import org.junit.runners.Parameterized; 056import org.junit.runners.Parameterized.Parameters; 057import org.slf4j.Logger; 058import org.slf4j.LoggerFactory; 059 060/** 061 * Test a multi-column scanner when there is a Bloom filter false-positive. This is needed for the 062 * multi-column Bloom filter optimization. 063 */ 064@RunWith(Parameterized.class) 065@Category({ RegionServerTests.class, SmallTests.class }) 066public class TestScanWithBloomError { 067 068 @ClassRule 069 public static final HBaseClassTestRule CLASS_RULE = 070 HBaseClassTestRule.forClass(TestScanWithBloomError.class); 071 072 private static final Logger LOG = LoggerFactory.getLogger(TestScanWithBloomError.class); 073 074 private static final String TABLE_NAME = "ScanWithBloomError"; 075 private static final String FAMILY = "myCF"; 076 private static final byte[] FAMILY_BYTES = Bytes.toBytes(FAMILY); 077 private static final String ROW = "theRow"; 078 private static final String QUALIFIER_PREFIX = "qual"; 079 private static final byte[] ROW_BYTES = Bytes.toBytes(ROW); 080 private static NavigableSet<Integer> allColIds = new TreeSet<>(); 081 private HRegion region; 082 private BloomType bloomType; 083 private FileSystem fs; 084 private Configuration conf; 085 086 private final static HBaseTestingUtility TEST_UTIL = HBaseTestingUtility.createLocalHTU(); 087 088 @Parameters 089 public static final Collection<Object[]> parameters() { 090 List<Object[]> configurations = new ArrayList<>(); 091 for (BloomType bloomType : BloomType.values()) { 092 configurations.add(new Object[] { bloomType }); 093 } 094 return configurations; 095 } 096 097 public TestScanWithBloomError(BloomType bloomType) { 098 this.bloomType = bloomType; 099 } 100 101 @Before 102 public void setUp() throws IOException { 103 conf = TEST_UTIL.getConfiguration(); 104 fs = FileSystem.get(conf); 105 conf.setInt(BloomFilterUtil.PREFIX_LENGTH_KEY, 10); 106 } 107 108 @Test 109 public void testThreeStoreFiles() throws IOException { 110 region = TEST_UTIL.createTestRegion(TABLE_NAME, 111 new HColumnDescriptor(FAMILY).setCompressionType(Compression.Algorithm.GZ) 112 .setBloomFilterType(bloomType).setMaxVersions(TestMultiColumnScanner.MAX_VERSIONS)); 113 createStoreFile(new int[] { 1, 2, 6 }); 114 createStoreFile(new int[] { 1, 2, 3, 7 }); 115 createStoreFile(new int[] { 1, 9 }); 116 scanColSet(new int[] { 1, 4, 6, 7 }, new int[] { 1, 6, 7 }); 117 118 HBaseTestingUtility.closeRegionAndWAL(region); 119 } 120 121 private void scanColSet(int[] colSet, int[] expectedResultCols) throws IOException { 122 LOG.info("Scanning column set: " + Arrays.toString(colSet)); 123 Scan scan = new Scan(ROW_BYTES, ROW_BYTES); 124 addColumnSetToScan(scan, colSet); 125 RegionScannerImpl scanner = region.getScanner(scan); 126 KeyValueHeap storeHeap = scanner.storeHeap; 127 assertEquals(0, storeHeap.getHeap().size()); 128 StoreScanner storeScanner = (StoreScanner) storeHeap.getCurrentForTesting(); 129 @SuppressWarnings({ "unchecked", "rawtypes" }) 130 List<StoreFileScanner> scanners = 131 (List<StoreFileScanner>) (List) storeScanner.getAllScannersForTesting(); 132 133 // Sort scanners by their HFile's modification time. 134 Collections.sort(scanners, new Comparator<StoreFileScanner>() { 135 @Override 136 public int compare(StoreFileScanner s1, StoreFileScanner s2) { 137 Path p1 = s1.getReader().getHFileReader().getPath(); 138 Path p2 = s2.getReader().getHFileReader().getPath(); 139 long t1, t2; 140 try { 141 t1 = fs.getFileStatus(p1).getModificationTime(); 142 t2 = fs.getFileStatus(p2).getModificationTime(); 143 } catch (IOException ex) { 144 throw new RuntimeException(ex); 145 } 146 return t1 < t2 ? -1 : t1 == t2 ? 1 : 0; 147 } 148 }); 149 150 StoreFileReader lastStoreFileReader = null; 151 for (StoreFileScanner sfScanner : scanners) 152 lastStoreFileReader = sfScanner.getReader(); 153 154 new HFilePrettyPrinter(conf).run( 155 new String[] { "-m", "-p", "-f", lastStoreFileReader.getHFileReader().getPath().toString() }); 156 157 // Disable Bloom filter for the last store file. The disabled Bloom filter 158 // will always return "true". 159 LOG.info("Disabling Bloom filter for: " + lastStoreFileReader.getHFileReader().getName()); 160 lastStoreFileReader.disableBloomFilterForTesting(); 161 162 List<Cell> allResults = new ArrayList<>(); 163 164 { // Limit the scope of results. 165 List<Cell> results = new ArrayList<>(); 166 while (scanner.next(results) || results.size() > 0) { 167 allResults.addAll(results); 168 results.clear(); 169 } 170 } 171 172 List<Integer> actualIds = new ArrayList<>(); 173 for (Cell kv : allResults) { 174 String qual = Bytes.toString(CellUtil.cloneQualifier(kv)); 175 assertTrue(qual.startsWith(QUALIFIER_PREFIX)); 176 actualIds.add(Integer.valueOf(qual.substring(QUALIFIER_PREFIX.length()))); 177 } 178 List<Integer> expectedIds = new ArrayList<>(); 179 for (int expectedId : expectedResultCols) 180 expectedIds.add(expectedId); 181 182 LOG.info("Column ids returned: " + actualIds + ", expected: " + expectedIds); 183 assertEquals(expectedIds.toString(), actualIds.toString()); 184 } 185 186 private void addColumnSetToScan(Scan scan, int[] colIds) { 187 for (int colId : colIds) { 188 scan.addColumn(FAMILY_BYTES, Bytes.toBytes(qualFromId(colId))); 189 } 190 } 191 192 private String qualFromId(int colId) { 193 return QUALIFIER_PREFIX + colId; 194 } 195 196 private void createStoreFile(int[] colIds) throws IOException { 197 Put p = new Put(ROW_BYTES); 198 for (int colId : colIds) { 199 long ts = Long.MAX_VALUE; 200 String qual = qualFromId(colId); 201 allColIds.add(colId); 202 KeyValue kv = KeyValueTestUtil.create(ROW, FAMILY, qual, ts, 203 TestMultiColumnScanner.createValue(ROW, qual, ts)); 204 p.add(kv); 205 } 206 region.put(p); 207 region.flush(true); 208 } 209 210}