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.mapreduce;
019
020import static org.junit.jupiter.api.Assertions.assertArrayEquals;
021import static org.junit.jupiter.api.Assertions.assertEquals;
022import static org.mockito.ArgumentMatchers.any;
023import static org.mockito.ArgumentMatchers.anyBoolean;
024import static org.mockito.Mockito.mock;
025import static org.mockito.Mockito.when;
026
027import java.io.IOException;
028import java.net.Inet6Address;
029import java.net.InetAddress;
030import java.net.UnknownHostException;
031import java.util.List;
032import java.util.Map;
033import java.util.TreeMap;
034import java.util.concurrent.ExecutorService;
035import org.apache.hadoop.conf.Configuration;
036import org.apache.hadoop.hbase.HBaseConfiguration;
037import org.apache.hadoop.hbase.HConstants;
038import org.apache.hadoop.hbase.HRegionLocation;
039import org.apache.hadoop.hbase.ServerName;
040import org.apache.hadoop.hbase.TableName;
041import org.apache.hadoop.hbase.client.Admin;
042import org.apache.hadoop.hbase.client.AsyncConnection;
043import org.apache.hadoop.hbase.client.BufferedMutator;
044import org.apache.hadoop.hbase.client.BufferedMutatorParams;
045import org.apache.hadoop.hbase.client.Connection;
046import org.apache.hadoop.hbase.client.ConnectionRegistry;
047import org.apache.hadoop.hbase.client.ConnectionUtils;
048import org.apache.hadoop.hbase.client.RegionInfo;
049import org.apache.hadoop.hbase.client.RegionInfoBuilder;
050import org.apache.hadoop.hbase.client.RegionLocator;
051import org.apache.hadoop.hbase.client.Table;
052import org.apache.hadoop.hbase.client.TableBuilder;
053import org.apache.hadoop.hbase.security.User;
054import org.apache.hadoop.hbase.testclassification.SmallTests;
055import org.apache.hadoop.hbase.util.Bytes;
056import org.apache.hadoop.hbase.util.Pair;
057import org.apache.hadoop.mapreduce.InputSplit;
058import org.apache.hadoop.mapreduce.JobContext;
059import org.junit.jupiter.api.Tag;
060import org.junit.jupiter.api.Test;
061import org.mockito.Mockito;
062import org.mockito.invocation.InvocationOnMock;
063import org.mockito.stubbing.Answer;
064
065@Tag(SmallTests.TAG)
066public class TestTableInputFormatBase {
067
068  @Test
069  public void testCreateNInputSplitsUniformPreservesOriginalBoundaries() throws IOException {
070    TableInputFormat inputFormat = new TableInputFormat();
071    for (byte[] startRow : new byte[][] { HConstants.EMPTY_START_ROW, Bytes.toBytes("start") }) {
072      TableSplit split =
073        new TableSplit(TableName.valueOf("test"), startRow, HConstants.EMPTY_END_ROW, "localhost");
074
075      List<InputSplit> splits = inputFormat.createNInputSplitsUniform(split, 2);
076
077      assertEquals(2, splits.size());
078      TableSplit first = (TableSplit) splits.get(0);
079      TableSplit last = (TableSplit) splits.get(1);
080      assertArrayEquals(startRow, first.getStartRow());
081      assertArrayEquals(first.getEndRow(), last.getStartRow());
082      assertArrayEquals(HConstants.EMPTY_END_ROW, last.getEndRow());
083    }
084  }
085
086  @Test
087  public void testReuseRegionSizeCalculator() throws IOException {
088    JobContext context = mock(JobContext.class);
089    Configuration conf = HBaseConfiguration.create();
090    conf.set(ConnectionUtils.HBASE_CLIENT_CONNECTION_IMPL,
091      ConnectionForMergeTesting.class.getName());
092    conf.set(TableInputFormat.INPUT_TABLE, "testTable");
093    conf.setBoolean(TableInputFormatBase.MAPREDUCE_INPUT_AUTOBALANCE, true);
094    when(context.getConfiguration()).thenReturn(conf);
095
096    TableInputFormat format = Mockito.spy(new TableInputFormatForMergeTesting());
097    format.setConf(conf);
098    // initialize so that table is set, otherwise cloneOnFinish
099    // will be true and each getSplits call will re-initialize everything
100    format.initialize(context);
101    format.getSplits(context);
102    format.getSplits(context);
103
104    // re-initialize which will cause the next getSplits call to create a new RegionSizeCalculator
105    format.initialize(context);
106    format.getSplits(context);
107    format.getSplits(context);
108
109    // should only be 2 despite calling getSplits 4 times
110    Mockito.verify(format, Mockito.times(2)).createRegionSizeCalculator(Mockito.any(),
111      Mockito.any());
112  }
113
114  @Test
115  public void testTableInputFormatBaseReverseDNSForIPv6() throws UnknownHostException {
116    String address = "ipv6.google.com";
117    String localhost = null;
118    InetAddress addr = null;
119    TableInputFormat inputFormat = new TableInputFormat();
120    try {
121      localhost = InetAddress.getByName(address).getCanonicalHostName();
122      addr = Inet6Address.getByName(address);
123    } catch (UnknownHostException e) {
124      // google.com is down, we can probably forgive this test.
125      return;
126    }
127    System.out.println("Should retrun the hostname for this host " + localhost + " addr : " + addr);
128    String actualHostName = inputFormat.reverseDNS(addr);
129    assertEquals("Should retrun the hostname for this host. Expected : " + localhost + " Actual : "
130      + actualHostName, localhost, actualHostName);
131  }
132
133  @Test
134  public void testNonSuccessiveSplitsAreNotMerged() throws IOException {
135    JobContext context = mock(JobContext.class);
136    Configuration conf = HBaseConfiguration.create();
137    conf.set(ConnectionUtils.HBASE_CLIENT_CONNECTION_IMPL,
138      ConnectionForMergeTesting.class.getName());
139    conf.set(TableInputFormat.INPUT_TABLE, "testTable");
140    conf.setBoolean(TableInputFormatBase.MAPREDUCE_INPUT_AUTOBALANCE, true);
141    when(context.getConfiguration()).thenReturn(conf);
142
143    TableInputFormat tifExclude = new TableInputFormatForMergeTesting();
144    tifExclude.setConf(conf);
145    // split["b", "c"] is excluded, split["o", "p"] and split["p", "q"] are merged,
146    // but split["a", "b"] and split["c", "d"] are not merged.
147    assertEquals(ConnectionForMergeTesting.START_KEYS.length - 1 - 1,
148      tifExclude.getSplits(context).size());
149  }
150
151  /**
152   * Subclass of {@link TableInputFormat} to use in {@link #testNonSuccessiveSplitsAreNotMerged}.
153   * This class overrides {@link TableInputFormatBase#includeRegionInSplit} to exclude specific
154   * splits.
155   */
156  private static class TableInputFormatForMergeTesting extends TableInputFormat {
157    private byte[] prefixStartKey = Bytes.toBytes("b");
158    private byte[] prefixEndKey = Bytes.toBytes("c");
159    private RegionSizeCalculator sizeCalculator;
160
161    /**
162     * Exclude regions which contain rows starting with "b".
163     */
164    @Override
165    protected boolean includeRegionInSplit(final byte[] startKey, final byte[] endKey) {
166      if (
167        Bytes.compareTo(startKey, prefixEndKey) < 0 && (Bytes.compareTo(prefixStartKey, endKey) < 0
168          || Bytes.equals(endKey, HConstants.EMPTY_END_ROW))
169      ) {
170        return false;
171      } else {
172        return true;
173      }
174    }
175
176    @Override
177    protected void initializeTable(Connection connection, TableName tableName) throws IOException {
178      super.initializeTable(connection, tableName);
179      ConnectionForMergeTesting cft = (ConnectionForMergeTesting) connection;
180      sizeCalculator = cft.getRegionSizeCalculator();
181    }
182
183    @Override
184    protected RegionSizeCalculator createRegionSizeCalculator(RegionLocator locator, Admin admin)
185      throws IOException {
186      return sizeCalculator;
187    }
188  }
189
190  /**
191   * Connection class to use in {@link #testNonSuccessiveSplitsAreNotMerged}. This class returns
192   * mocked {@link Table}, {@link RegionLocator}, {@link RegionSizeCalculator}, and {@link Admin}.
193   */
194  private static class ConnectionForMergeTesting implements Connection {
195    public static final byte[][] SPLITS = new byte[][] { Bytes.toBytes("a"), Bytes.toBytes("b"),
196      Bytes.toBytes("c"), Bytes.toBytes("d"), Bytes.toBytes("e"), Bytes.toBytes("f"),
197      Bytes.toBytes("g"), Bytes.toBytes("h"), Bytes.toBytes("i"), Bytes.toBytes("j"),
198      Bytes.toBytes("k"), Bytes.toBytes("l"), Bytes.toBytes("m"), Bytes.toBytes("n"),
199      Bytes.toBytes("o"), Bytes.toBytes("p"), Bytes.toBytes("q"), Bytes.toBytes("r"),
200      Bytes.toBytes("s"), Bytes.toBytes("t"), Bytes.toBytes("u"), Bytes.toBytes("v"),
201      Bytes.toBytes("w"), Bytes.toBytes("x"), Bytes.toBytes("y"), Bytes.toBytes("z") };
202
203    public static final byte[][] START_KEYS;
204    public static final byte[][] END_KEYS;
205    static {
206      START_KEYS = new byte[SPLITS.length + 1][];
207      START_KEYS[0] = HConstants.EMPTY_BYTE_ARRAY;
208      for (int i = 0; i < SPLITS.length; i++) {
209        START_KEYS[i + 1] = SPLITS[i];
210      }
211
212      END_KEYS = new byte[SPLITS.length + 1][];
213      for (int i = 0; i < SPLITS.length; i++) {
214        END_KEYS[i] = SPLITS[i];
215      }
216      END_KEYS[SPLITS.length] = HConstants.EMPTY_BYTE_ARRAY;
217    }
218
219    public static final Map<byte[], Long> SIZE_MAP = new TreeMap<>(Bytes.BYTES_COMPARATOR);
220    static {
221      for (byte[] startKey : START_KEYS) {
222        SIZE_MAP.put(startKey, 1024L * 1024L * 1024L);
223      }
224      SIZE_MAP.put(Bytes.toBytes("a"), 200L * 1024L * 1024L);
225      SIZE_MAP.put(Bytes.toBytes("b"), 200L * 1024L * 1024L);
226      SIZE_MAP.put(Bytes.toBytes("c"), 200L * 1024L * 1024L);
227      SIZE_MAP.put(Bytes.toBytes("o"), 200L * 1024L * 1024L);
228      SIZE_MAP.put(Bytes.toBytes("p"), 200L * 1024L * 1024L);
229    }
230
231    ConnectionForMergeTesting(Configuration conf, ExecutorService pool, User user,
232      ConnectionRegistry registry, Map<String, byte[]> connectionAttributes) throws IOException {
233    }
234
235    @Override
236    public void abort(String why, Throwable e) {
237    }
238
239    @Override
240    public boolean isAborted() {
241      return false;
242    }
243
244    @Override
245    public Configuration getConfiguration() {
246      throw new UnsupportedOperationException();
247    }
248
249    @Override
250    public Table getTable(TableName tableName) throws IOException {
251      Table table = mock(Table.class);
252      when(table.getName()).thenReturn(tableName);
253      return table;
254    }
255
256    @Override
257    public Table getTable(TableName tableName, ExecutorService pool) throws IOException {
258      throw new UnsupportedOperationException();
259    }
260
261    @Override
262    public BufferedMutator getBufferedMutator(TableName tableName) throws IOException {
263      throw new UnsupportedOperationException();
264    }
265
266    @Override
267    public BufferedMutator getBufferedMutator(BufferedMutatorParams params) throws IOException {
268      throw new UnsupportedOperationException();
269    }
270
271    @Override
272    public RegionLocator getRegionLocator(TableName tableName) throws IOException {
273      final Map<byte[], HRegionLocation> locationMap = new TreeMap<>(Bytes.BYTES_COMPARATOR);
274      for (byte[] startKey : START_KEYS) {
275        HRegionLocation hrl =
276          new HRegionLocation(RegionInfoBuilder.newBuilder(tableName).setStartKey(startKey).build(),
277            ServerName.valueOf("localhost", 0, 0));
278        locationMap.put(startKey, hrl);
279      }
280
281      RegionLocator locator = mock(RegionLocator.class);
282      when(locator.getRegionLocation(any(byte[].class), anyBoolean()))
283        .thenAnswer(new Answer<HRegionLocation>() {
284          @Override
285          public HRegionLocation answer(InvocationOnMock invocationOnMock) throws Throwable {
286            Object[] args = invocationOnMock.getArguments();
287            byte[] key = (byte[]) args[0];
288            return locationMap.get(key);
289          }
290        });
291      when(locator.getStartEndKeys())
292        .thenReturn(new Pair<byte[][], byte[][]>(START_KEYS, END_KEYS));
293      return locator;
294    }
295
296    public RegionSizeCalculator getRegionSizeCalculator() {
297      RegionSizeCalculator sizeCalculator = mock(RegionSizeCalculator.class);
298      when(sizeCalculator.getRegionSize(any(byte[].class))).thenAnswer(new Answer<Long>() {
299        @Override
300        public Long answer(InvocationOnMock invocationOnMock) throws Throwable {
301          Object[] args = invocationOnMock.getArguments();
302          byte[] regionId = (byte[]) args[0];
303          byte[] startKey = RegionInfo.getStartKey(regionId);
304          return SIZE_MAP.get(startKey);
305        }
306      });
307      return sizeCalculator;
308    }
309
310    @Override
311    public Admin getAdmin() throws IOException {
312      Admin admin = mock(Admin.class);
313      // return non-null admin to pass null checks
314      return admin;
315    }
316
317    @Override
318    public void close() throws IOException {
319    }
320
321    @Override
322    public boolean isClosed() {
323      return false;
324    }
325
326    @Override
327    public TableBuilder getTableBuilder(TableName tableName, ExecutorService pool) {
328      throw new UnsupportedOperationException();
329    }
330
331    @Override
332    public void clearRegionLocationCache() {
333    }
334
335    @Override
336    public AsyncConnection toAsyncConnection() {
337      throw new UnsupportedOperationException();
338    }
339
340    @Override
341    public String getClusterId() {
342      return null;
343    }
344  }
345}