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.replication; 019 020import static org.junit.Assert.assertArrayEquals; 021import static org.junit.Assert.assertEquals; 022import static org.junit.Assert.fail; 023 024import java.io.IOException; 025import java.util.ArrayList; 026import java.util.List; 027import java.util.concurrent.atomic.AtomicInteger; 028import java.util.stream.Collectors; 029import org.apache.hadoop.conf.Configuration; 030import org.apache.hadoop.fs.FileSystem; 031import org.apache.hadoop.fs.Path; 032import org.apache.hadoop.hbase.HBaseConfiguration; 033import org.apache.hadoop.hbase.HBaseTestingUtil; 034import org.apache.hadoop.hbase.HConstants; 035import org.apache.hadoop.hbase.ServerName; 036import org.apache.hadoop.hbase.TableName; 037import org.apache.hadoop.hbase.client.Admin; 038import org.apache.hadoop.hbase.client.ColumnFamilyDescriptorBuilder; 039import org.apache.hadoop.hbase.client.Connection; 040import org.apache.hadoop.hbase.client.ConnectionFactory; 041import org.apache.hadoop.hbase.client.Delete; 042import org.apache.hadoop.hbase.client.Get; 043import org.apache.hadoop.hbase.client.Put; 044import org.apache.hadoop.hbase.client.Result; 045import org.apache.hadoop.hbase.client.ResultScanner; 046import org.apache.hadoop.hbase.client.Scan; 047import org.apache.hadoop.hbase.client.Table; 048import org.apache.hadoop.hbase.client.TableDescriptor; 049import org.apache.hadoop.hbase.client.TableDescriptorBuilder; 050import org.apache.hadoop.hbase.replication.regionserver.HBaseInterClusterReplicationEndpoint; 051import org.apache.hadoop.hbase.util.Bytes; 052import org.apache.hadoop.hbase.util.JVMClusterUtil; 053import org.apache.hadoop.hbase.wal.WAL; 054import org.apache.hadoop.hbase.zookeeper.MiniZooKeeperCluster; 055import org.junit.After; 056import org.junit.AfterClass; 057import org.junit.Before; 058import org.junit.BeforeClass; 059import org.slf4j.Logger; 060import org.slf4j.LoggerFactory; 061 062import org.apache.hbase.thirdparty.com.google.common.collect.ImmutableList; 063import org.apache.hbase.thirdparty.com.google.common.collect.ImmutableMap; 064import org.apache.hbase.thirdparty.com.google.common.collect.Lists; 065import org.apache.hbase.thirdparty.com.google.common.io.Closeables; 066 067/** 068 * This class is only a base for other integration-level replication tests. Do not add tests here. 069 * TestReplicationSmallTests is where tests that don't require bring machines up/down should go All 070 * other tests should have their own classes and extend this one 071 */ 072public class TestReplicationBase { 073 private static final Logger LOG = LoggerFactory.getLogger(TestReplicationBase.class); 074 protected static Connection connection1; 075 protected static Connection connection2; 076 protected static Configuration CONF_WITH_LOCALFS; 077 078 protected static Admin hbaseAdmin; 079 080 protected static Table htable1; 081 protected static Table htable2; 082 083 protected static final HBaseTestingUtil UTIL1 = new HBaseTestingUtil(); 084 protected static final HBaseTestingUtil UTIL2 = new HBaseTestingUtil(); 085 protected static Configuration CONF1 = UTIL1.getConfiguration(); 086 protected static Configuration CONF2 = UTIL2.getConfiguration(); 087 088 protected static int NUM_SLAVES1 = 1; 089 protected static int NUM_SLAVES2 = 1; 090 protected static final int NB_ROWS_IN_BATCH = 100; 091 protected static final int NB_ROWS_IN_BIG_BATCH = NB_ROWS_IN_BATCH * 10; 092 protected static final long SLEEP_TIME = 500; 093 protected static final int NB_RETRIES = 50; 094 protected static AtomicInteger replicateCount = new AtomicInteger(); 095 protected static volatile List<WAL.Entry> replicatedEntries = Lists.newArrayList(); 096 097 protected static final TableName tableName = TableName.valueOf("test"); 098 protected static final byte[] famName = Bytes.toBytes("f"); 099 protected static final byte[] row = Bytes.toBytes("row"); 100 protected static final byte[] noRepfamName = Bytes.toBytes("norep"); 101 protected static final String PEER_ID2 = "2"; 102 103 protected boolean isSerialPeer() { 104 return false; 105 } 106 107 protected boolean isSyncPeer() { 108 return false; 109 } 110 111 protected final void cleanUp() throws IOException, InterruptedException { 112 // Starting and stopping replication can make us miss new logs, 113 // rolling like this makes sure the most recent one gets added to the queue 114 for (JVMClusterUtil.RegionServerThread r : UTIL1.getHBaseCluster().getRegionServerThreads()) { 115 UTIL1.getAdmin().rollWALWriter(r.getRegionServer().getServerName()); 116 } 117 int rowCount = UTIL1.countRows(tableName); 118 UTIL1.deleteTableData(tableName); 119 // truncating the table will send one Delete per row to the slave cluster 120 // in an async fashion, which is why we cannot just call deleteTableData on 121 // utility2 since late writes could make it to the slave in some way. 122 // Instead, we truncate the first table and wait for all the Deletes to 123 // make it to the slave. 124 Scan scan = new Scan(); 125 int lastCount = 0; 126 for (int i = 0; i < NB_RETRIES; i++) { 127 if (i == NB_RETRIES - 1) { 128 fail("Waited too much time for truncate"); 129 } 130 ResultScanner scanner = htable2.getScanner(scan); 131 Result[] res = scanner.next(rowCount); 132 scanner.close(); 133 if (res.length != 0) { 134 if (res.length < lastCount) { 135 i--; // Don't increment timeout if we make progress 136 } 137 lastCount = res.length; 138 LOG.info("Still got " + res.length + " rows"); 139 Thread.sleep(SLEEP_TIME); 140 } else { 141 break; 142 } 143 } 144 } 145 146 protected static void waitForReplication(int expectedRows, int retries) 147 throws IOException, InterruptedException { 148 waitForReplication(htable2, expectedRows, retries); 149 } 150 151 protected static void waitForReplication(Table table, int expectedRows, int retries) 152 throws IOException, InterruptedException { 153 Scan scan; 154 for (int i = 0; i < retries; i++) { 155 scan = new Scan(); 156 if (i == retries - 1) { 157 fail("Waited too much time for normal batch replication"); 158 } 159 int count = 0; 160 try (ResultScanner scanner = table.getScanner(scan)) { 161 while (scanner.next() != null) { 162 count++; 163 } 164 } 165 if (count != expectedRows) { 166 LOG.info("Only got " + count + " rows"); 167 Thread.sleep(SLEEP_TIME); 168 } else { 169 break; 170 } 171 } 172 } 173 174 protected static void loadData(String prefix, byte[] row) throws IOException { 175 loadData(prefix, row, famName); 176 } 177 178 protected static void loadData(String prefix, byte[] row, byte[] familyName) throws IOException { 179 List<Put> puts = new ArrayList<>(NB_ROWS_IN_BATCH); 180 for (int i = 0; i < NB_ROWS_IN_BATCH; i++) { 181 Put put = new Put(Bytes.toBytes(prefix + Integer.toString(i))); 182 put.addColumn(familyName, row, row); 183 puts.add(put); 184 } 185 htable1.put(puts); 186 } 187 188 protected static void setupConfig(HBaseTestingUtil util, String znodeParent) { 189 Configuration conf = util.getConfiguration(); 190 conf.set(HConstants.ZOOKEEPER_ZNODE_PARENT, znodeParent); 191 // We don't want too many edits per batch sent to the ReplicationEndpoint to trigger 192 // sufficient number of events. But we don't want to go too low because 193 // HBaseInterClusterReplicationEndpoint partitions entries into batches and we want 194 // more than one batch sent to the peer cluster for better testing. 195 conf.setInt("replication.source.size.capacity", 102400); 196 conf.setLong("replication.source.sleepforretries", 100); 197 conf.setInt("hbase.regionserver.maxlogs", 10); 198 conf.setLong("hbase.master.logcleaner.ttl", 10); 199 conf.setInt("zookeeper.recovery.retry", 1); 200 conf.setInt("zookeeper.recovery.retry.intervalmill", 10); 201 conf.setLong(HConstants.THREAD_WAKE_FREQUENCY, 100); 202 conf.setInt("replication.stats.thread.period.seconds", 5); 203 conf.setBoolean("hbase.tests.use.shortcircuit.reads", false); 204 conf.setLong("replication.sleep.before.failover", 2000); 205 conf.setInt("replication.source.maxretriesmultiplier", 10); 206 conf.setFloat("replication.source.ratio", 1.0f); 207 conf.setBoolean("replication.source.eof.autorecovery", true); 208 conf.setLong("hbase.serial.replication.waiting.ms", 100); 209 } 210 211 static void configureClusters(HBaseTestingUtil util1, HBaseTestingUtil util2) { 212 setupConfig(util1, "/1"); 213 setupConfig(util2, "/2"); 214 215 Configuration conf2 = util2.getConfiguration(); 216 conf2.set(HConstants.ZOOKEEPER_ZNODE_PARENT, "/2"); 217 conf2.setInt(HConstants.HBASE_CLIENT_RETRIES_NUMBER, 6); 218 conf2.setBoolean("hbase.tests.use.shortcircuit.reads", false); 219 } 220 221 protected static void restartSourceCluster(int numSlaves) throws Exception { 222 Closeables.close(hbaseAdmin, true); 223 Closeables.close(htable1, true); 224 UTIL1.shutdownMiniHBaseCluster(); 225 UTIL1.restartHBaseCluster(numSlaves); 226 // Invalidate the cached connection state. 227 CONF1 = UTIL1.getConfiguration(); 228 hbaseAdmin = UTIL1.getAdmin(); 229 Connection connection1 = UTIL1.getConnection(); 230 htable1 = connection1.getTable(tableName); 231 } 232 233 static void restartTargetHBaseCluster(int numSlaves) throws Exception { 234 Closeables.close(htable2, true); 235 UTIL2.restartHBaseCluster(numSlaves); 236 // Invalidate the cached connection state 237 CONF2 = UTIL2.getConfiguration(); 238 htable2 = UTIL2.getConnection().getTable(tableName); 239 } 240 241 protected static void createTable(TableName tableName) throws IOException { 242 TableDescriptor table = TableDescriptorBuilder.newBuilder(tableName) 243 .setColumnFamily(ColumnFamilyDescriptorBuilder.newBuilder(famName).setMaxVersions(100) 244 .setScope(HConstants.REPLICATION_SCOPE_GLOBAL).build()) 245 .setColumnFamily(ColumnFamilyDescriptorBuilder.of(noRepfamName)).build(); 246 UTIL1.createTable(table, HBaseTestingUtil.KEYS_FOR_HBA_CREATE_TABLE); 247 UTIL2.createTable(table, HBaseTestingUtil.KEYS_FOR_HBA_CREATE_TABLE); 248 UTIL1.waitUntilAllRegionsAssigned(tableName); 249 UTIL2.waitUntilAllRegionsAssigned(tableName); 250 } 251 252 private static void startClusters() throws Exception { 253 UTIL1.startMiniZKCluster(); 254 MiniZooKeeperCluster miniZK = UTIL1.getZkCluster(); 255 LOG.info("Setup first Zk"); 256 257 UTIL2.setZkCluster(miniZK); 258 LOG.info("Setup second Zk"); 259 260 CONF_WITH_LOCALFS = HBaseConfiguration.create(CONF1); 261 UTIL1.startMiniCluster(NUM_SLAVES1); 262 // Have a bunch of slave servers, because inter-cluster shipping logic uses number of sinks 263 // as a component in deciding maximum number of parallel batches to send to the peer cluster. 264 UTIL2.startMiniCluster(NUM_SLAVES2); 265 266 connection1 = ConnectionFactory.createConnection(CONF1); 267 connection2 = ConnectionFactory.createConnection(CONF2); 268 hbaseAdmin = connection1.getAdmin(); 269 270 createTable(tableName); 271 htable1 = connection1.getTable(tableName); 272 htable2 = connection2.getTable(tableName); 273 } 274 275 @BeforeClass 276 public static void setUpBeforeClass() throws Exception { 277 configureClusters(UTIL1, UTIL2); 278 startClusters(); 279 } 280 281 private boolean peerExist(String peerId) throws IOException { 282 return hbaseAdmin.listReplicationPeers().stream().anyMatch(p -> peerId.equals(p.getPeerId())); 283 } 284 285 // can be override in tests, in case you need to use zk based uri, or the old style uri 286 protected String getClusterKey(HBaseTestingUtil util) throws Exception { 287 return util.getRpcConnnectionURI(); 288 } 289 290 protected final void addPeer(String peerId, TableName tableName) throws Exception { 291 if (!peerExist(peerId)) { 292 ReplicationPeerConfigBuilder builder = ReplicationPeerConfig.newBuilder() 293 .setClusterKey(getClusterKey(UTIL2)).setSerial(isSerialPeer()) 294 .setReplicationEndpointImpl(ReplicationEndpointTest.class.getName()); 295 if (isSyncPeer()) { 296 FileSystem fs2 = UTIL2.getTestFileSystem(); 297 // The remote wal dir is not important as we do not use it in DA state, here we only need to 298 // confirm that a sync peer in DA state can still replicate data to remote cluster 299 // asynchronously. 300 builder.setReplicateAllUserTables(false) 301 .setTableCFsMap(ImmutableMap.of(tableName, ImmutableList.of())) 302 .setRemoteWALDir(new Path("/RemoteWAL") 303 .makeQualified(fs2.getUri(), fs2.getWorkingDirectory()).toUri().toString()); 304 } 305 hbaseAdmin.addReplicationPeer(peerId, builder.build()); 306 } 307 } 308 309 @Before 310 public void setUpBase() throws Exception { 311 addPeer(PEER_ID2, tableName); 312 } 313 314 protected final void removePeer(String peerId) throws Exception { 315 if (peerExist(peerId)) { 316 hbaseAdmin.removeReplicationPeer(peerId); 317 } 318 } 319 320 @After 321 public void tearDownBase() throws Exception { 322 removePeer(PEER_ID2); 323 } 324 325 protected static void runSimplePutDeleteTest() throws IOException, InterruptedException { 326 Put put = new Put(row); 327 put.addColumn(famName, row, row); 328 329 htable1 = UTIL1.getConnection().getTable(tableName); 330 htable1.put(put); 331 332 Get get = new Get(row); 333 for (int i = 0; i < NB_RETRIES; i++) { 334 if (i == NB_RETRIES - 1) { 335 fail("Waited too much time for put replication"); 336 } 337 Result res = htable2.get(get); 338 if (res.isEmpty()) { 339 LOG.info("Row not available"); 340 Thread.sleep(SLEEP_TIME); 341 } else { 342 assertArrayEquals(row, res.value()); 343 break; 344 } 345 } 346 347 Delete del = new Delete(row); 348 htable1.delete(del); 349 350 get = new Get(row); 351 for (int i = 0; i < NB_RETRIES; i++) { 352 if (i == NB_RETRIES - 1) { 353 fail("Waited too much time for del replication"); 354 } 355 Result res = htable2.get(get); 356 if (res.size() >= 1) { 357 LOG.info("Row not deleted"); 358 Thread.sleep(SLEEP_TIME); 359 } else { 360 break; 361 } 362 } 363 } 364 365 protected static void runSmallBatchTest() throws IOException, InterruptedException { 366 // normal Batch tests 367 loadData("", row); 368 369 Scan scan = new Scan(); 370 371 ResultScanner scanner1 = htable1.getScanner(scan); 372 Result[] res1 = scanner1.next(NB_ROWS_IN_BATCH); 373 scanner1.close(); 374 assertEquals(NB_ROWS_IN_BATCH, res1.length); 375 376 waitForReplication(NB_ROWS_IN_BATCH, NB_RETRIES); 377 } 378 379 protected static void stopAllRegionServers(HBaseTestingUtil util) throws IOException { 380 List<ServerName> rses = util.getMiniHBaseCluster().getRegionServerThreads().stream() 381 .map(t -> t.getRegionServer().getServerName()).collect(Collectors.toList()); 382 for (ServerName rs : rses) { 383 util.getMiniHBaseCluster().stopRegionServer(rs); 384 } 385 } 386 387 @AfterClass 388 public static void tearDownAfterClass() throws Exception { 389 if (htable2 != null) { 390 htable2.close(); 391 } 392 if (htable1 != null) { 393 htable1.close(); 394 } 395 if (hbaseAdmin != null) { 396 hbaseAdmin.close(); 397 } 398 399 if (connection2 != null) { 400 connection2.close(); 401 } 402 if (connection1 != null) { 403 connection1.close(); 404 } 405 UTIL2.shutdownMiniCluster(); 406 UTIL1.shutdownMiniCluster(); 407 } 408 409 /** 410 * Custom replication endpoint to keep track of replication status for tests. 411 */ 412 public static class ReplicationEndpointTest extends HBaseInterClusterReplicationEndpoint { 413 public ReplicationEndpointTest() { 414 replicateCount.set(0); 415 } 416 417 @Override 418 public boolean replicate(ReplicateContext replicateContext) { 419 replicateCount.incrementAndGet(); 420 replicatedEntries.addAll(replicateContext.getEntries()); 421 422 return super.replicate(replicateContext); 423 } 424 } 425}