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.wal; 019 020import static org.hamcrest.CoreMatchers.containsString; 021import static org.hamcrest.MatcherAssert.assertThat; 022import static org.junit.Assert.assertEquals; 023import static org.junit.Assert.assertNotEquals; 024import static org.junit.Assert.assertNotNull; 025import static org.junit.Assert.assertNull; 026import static org.junit.Assert.assertTrue; 027import static org.junit.Assert.fail; 028import static org.mockito.Mockito.mock; 029import static org.mockito.Mockito.when; 030 031import java.io.IOException; 032import java.util.ArrayList; 033import java.util.Collection; 034import java.util.Collections; 035import java.util.Comparator; 036import java.util.HashSet; 037import java.util.List; 038import java.util.Map; 039import java.util.NavigableMap; 040import java.util.Set; 041import java.util.TreeMap; 042import java.util.UUID; 043import java.util.concurrent.ConcurrentSkipListMap; 044import java.util.concurrent.CountDownLatch; 045import java.util.concurrent.ExecutorService; 046import java.util.concurrent.Executors; 047import java.util.concurrent.atomic.AtomicBoolean; 048import java.util.stream.Collectors; 049import org.apache.hadoop.conf.Configuration; 050import org.apache.hadoop.fs.FileStatus; 051import org.apache.hadoop.fs.FileSystem; 052import org.apache.hadoop.fs.Path; 053import org.apache.hadoop.hbase.Coprocessor; 054import org.apache.hadoop.hbase.ExtendedCellScanner; 055import org.apache.hadoop.hbase.HBaseConfiguration; 056import org.apache.hadoop.hbase.HBaseTestingUtil; 057import org.apache.hadoop.hbase.HConstants; 058import org.apache.hadoop.hbase.KeyValue; 059import org.apache.hadoop.hbase.ServerName; 060import org.apache.hadoop.hbase.TableName; 061import org.apache.hadoop.hbase.client.ColumnFamilyDescriptor; 062import org.apache.hadoop.hbase.client.ColumnFamilyDescriptorBuilder; 063import org.apache.hadoop.hbase.client.Durability; 064import org.apache.hadoop.hbase.client.Get; 065import org.apache.hadoop.hbase.client.Put; 066import org.apache.hadoop.hbase.client.RegionInfo; 067import org.apache.hadoop.hbase.client.RegionInfoBuilder; 068import org.apache.hadoop.hbase.client.Result; 069import org.apache.hadoop.hbase.client.TableDescriptor; 070import org.apache.hadoop.hbase.client.TableDescriptorBuilder; 071import org.apache.hadoop.hbase.coprocessor.CoprocessorHost; 072import org.apache.hadoop.hbase.coprocessor.SampleRegionWALCoprocessor; 073import org.apache.hadoop.hbase.regionserver.ChunkCreator; 074import org.apache.hadoop.hbase.regionserver.FlushPolicy; 075import org.apache.hadoop.hbase.regionserver.FlushPolicyFactory; 076import org.apache.hadoop.hbase.regionserver.HRegion; 077import org.apache.hadoop.hbase.regionserver.HStore; 078import org.apache.hadoop.hbase.regionserver.MemStoreLAB; 079import org.apache.hadoop.hbase.regionserver.MultiVersionConcurrencyControl; 080import org.apache.hadoop.hbase.regionserver.RegionServerServices; 081import org.apache.hadoop.hbase.regionserver.SequenceId; 082import org.apache.hadoop.hbase.util.Bytes; 083import org.apache.hadoop.hbase.util.CommonFSUtils; 084import org.apache.hadoop.hbase.util.EnvironmentEdge; 085import org.apache.hadoop.hbase.util.EnvironmentEdgeManager; 086import org.apache.hadoop.hbase.util.Threads; 087import org.apache.hadoop.hbase.wal.AbstractFSWALProvider; 088import org.apache.hadoop.hbase.wal.WAL; 089import org.apache.hadoop.hbase.wal.WALEdit; 090import org.apache.hadoop.hbase.wal.WALEditInternalHelper; 091import org.apache.hadoop.hbase.wal.WALKey; 092import org.apache.hadoop.hbase.wal.WALKeyImpl; 093import org.junit.AfterClass; 094import org.junit.Before; 095import org.junit.BeforeClass; 096import org.junit.Rule; 097import org.junit.Test; 098import org.junit.rules.TestName; 099import org.slf4j.Logger; 100import org.slf4j.LoggerFactory; 101 102public abstract class AbstractTestFSWAL { 103 104 protected static final Logger LOG = LoggerFactory.getLogger(AbstractTestFSWAL.class); 105 106 protected static Configuration CONF; 107 protected static FileSystem FS; 108 protected static Path DIR; 109 protected final static HBaseTestingUtil TEST_UTIL = new HBaseTestingUtil(); 110 111 @Rule 112 public final TestName currentTest = new TestName(); 113 114 @Before 115 public void setUp() throws Exception { 116 FileStatus[] entries = FS.listStatus(new Path("/")); 117 for (FileStatus dir : entries) { 118 FS.delete(dir.getPath(), true); 119 } 120 final Path hbaseDir = TEST_UTIL.createRootDir(); 121 final Path hbaseWALDir = TEST_UTIL.createWALRootDir(); 122 DIR = new Path(hbaseWALDir, currentTest.getMethodName()); 123 assertNotEquals(hbaseDir, hbaseWALDir); 124 FS.mkdirs(DIR); 125 } 126 127 @BeforeClass 128 public static void setUpBeforeClass() throws Exception { 129 // Make block sizes small. 130 TEST_UTIL.getConfiguration().setInt("dfs.blocksize", 1024 * 1024); 131 // quicker heartbeat interval for faster DN death notification 132 TEST_UTIL.getConfiguration().setInt("dfs.namenode.heartbeat.recheck-interval", 5000); 133 TEST_UTIL.getConfiguration().setInt("dfs.heartbeat.interval", 1); 134 TEST_UTIL.getConfiguration().setInt("dfs.client.socket-timeout", 5000); 135 136 // faster failover with cluster.shutdown();fs.close() idiom 137 TEST_UTIL.getConfiguration().setInt("hbase.ipc.client.connect.max.retries", 1); 138 TEST_UTIL.getConfiguration().setInt("dfs.client.block.recovery.retries", 1); 139 TEST_UTIL.getConfiguration().setInt("hbase.ipc.client.connection.maxidletime", 500); 140 TEST_UTIL.getConfiguration().set(CoprocessorHost.WAL_COPROCESSOR_CONF_KEY, 141 SampleRegionWALCoprocessor.class.getName()); 142 TEST_UTIL.startMiniDFSCluster(3); 143 144 CONF = TEST_UTIL.getConfiguration(); 145 FS = TEST_UTIL.getDFSCluster().getFileSystem(); 146 } 147 148 @AfterClass 149 public static void tearDownAfterClass() throws Exception { 150 TEST_UTIL.shutdownMiniCluster(); 151 } 152 153 protected abstract AbstractFSWAL<?> newWAL(FileSystem fs, Path rootDir, String WALDir, 154 String archiveDir, Configuration conf, List<WALActionsListener> listeners, 155 boolean failIfWALExists, String prefix, String suffix) throws IOException; 156 157 protected abstract AbstractFSWAL<?> newSlowWAL(FileSystem fs, Path rootDir, String WALDir, 158 String archiveDir, Configuration conf, List<WALActionsListener> listeners, 159 boolean failIfWALExists, String prefix, String suffix, Runnable action) throws IOException; 160 161 /** 162 * A loaded WAL coprocessor won't break existing WAL test cases. 163 */ 164 @Test 165 public void testWALCoprocessorLoaded() throws Exception { 166 // test to see whether the coprocessor is loaded or not. 167 AbstractFSWAL<?> wal = null; 168 try { 169 wal = newWAL(FS, CommonFSUtils.getWALRootDir(CONF), DIR.toString(), 170 HConstants.HREGION_OLDLOGDIR_NAME, CONF, null, true, null, null); 171 WALCoprocessorHost host = wal.getCoprocessorHost(); 172 Coprocessor c = host.findCoprocessor(SampleRegionWALCoprocessor.class); 173 assertNotNull(c); 174 } finally { 175 if (wal != null) { 176 wal.close(); 177 } 178 } 179 } 180 181 protected void addEdits(WAL log, RegionInfo hri, TableDescriptor htd, int times, 182 MultiVersionConcurrencyControl mvcc, NavigableMap<byte[], Integer> scopes, String cf) 183 throws IOException { 184 final byte[] row = Bytes.toBytes(cf); 185 for (int i = 0; i < times; i++) { 186 long timestamp = EnvironmentEdgeManager.currentTime(); 187 WALEdit cols = new WALEdit(); 188 WALEditInternalHelper.addExtendedCell(cols, new KeyValue(row, row, row, timestamp, row)); 189 WALKeyImpl key = 190 new WALKeyImpl(hri.getEncodedNameAsBytes(), htd.getTableName(), SequenceId.NO_SEQUENCE_ID, 191 timestamp, WALKey.EMPTY_UUIDS, HConstants.NO_NONCE, HConstants.NO_NONCE, mvcc, scopes); 192 log.appendData(hri, key, cols); 193 } 194 log.sync(); 195 } 196 197 /** 198 * helper method to simulate region flush for a WAL. 199 */ 200 protected void flushRegion(WAL wal, byte[] regionEncodedName, Set<byte[]> flushedFamilyNames) { 201 wal.startCacheFlush(regionEncodedName, flushedFamilyNames); 202 wal.completeCacheFlush(regionEncodedName, HConstants.NO_SEQNUM); 203 } 204 205 /** 206 * tests the log comparator. Ensure that we are not mixing meta logs with non-meta logs (throws 207 * exception if we do). Comparison is based on the timestamp present in the wal name. 208 */ 209 @Test 210 public void testWALComparator() throws Exception { 211 AbstractFSWAL<?> wal1 = null; 212 AbstractFSWAL<?> walMeta = null; 213 try { 214 wal1 = newWAL(FS, CommonFSUtils.getWALRootDir(CONF), DIR.toString(), 215 HConstants.HREGION_OLDLOGDIR_NAME, CONF, null, true, null, null); 216 LOG.debug("Log obtained is: " + wal1); 217 Comparator<Path> comp = wal1.LOG_NAME_COMPARATOR; 218 Path p1 = wal1.computeFilename(11); 219 Path p2 = wal1.computeFilename(12); 220 // comparing with itself returns 0 221 assertTrue(comp.compare(p1, p1) == 0); 222 // comparing with different filenum. 223 assertTrue(comp.compare(p1, p2) < 0); 224 walMeta = newWAL(FS, CommonFSUtils.getWALRootDir(CONF), DIR.toString(), 225 HConstants.HREGION_OLDLOGDIR_NAME, CONF, null, true, null, 226 AbstractFSWALProvider.META_WAL_PROVIDER_ID); 227 Comparator<Path> compMeta = walMeta.LOG_NAME_COMPARATOR; 228 229 Path p1WithMeta = walMeta.computeFilename(11); 230 Path p2WithMeta = walMeta.computeFilename(12); 231 assertTrue(compMeta.compare(p1WithMeta, p1WithMeta) == 0); 232 assertTrue(compMeta.compare(p1WithMeta, p2WithMeta) < 0); 233 // mixing meta and non-meta logs gives error 234 boolean ex = false; 235 try { 236 comp.compare(p1WithMeta, p2); 237 } catch (IllegalArgumentException e) { 238 ex = true; 239 } 240 assertTrue("Comparator doesn't complain while checking meta log files", ex); 241 boolean exMeta = false; 242 try { 243 compMeta.compare(p1WithMeta, p2); 244 } catch (IllegalArgumentException e) { 245 exMeta = true; 246 } 247 assertTrue("Meta comparator doesn't complain while checking log files", exMeta); 248 } finally { 249 if (wal1 != null) { 250 wal1.close(); 251 } 252 if (walMeta != null) { 253 walMeta.close(); 254 } 255 } 256 } 257 258 // now we will close asynchronously and will not archive a wal file unless it is fully closed, so 259 // sometimes we need to wait a bit before asserting, especially when you want to test the removal 260 // of numRolledLogFiles 261 private void waitNumRolledLogFiles(AbstractFSWAL<?> wal, int expected) { 262 TEST_UTIL.waitFor(5000, () -> wal.getNumRolledLogFiles() == expected); 263 } 264 265 private void testFindMemStoresEligibleForFlush(AbstractFSWAL<?> wal) throws IOException { 266 String cf1 = "cf1"; 267 String cf2 = "cf2"; 268 String cf3 = "cf3"; 269 TableDescriptor t1 = TableDescriptorBuilder.newBuilder(TableName.valueOf("t1")) 270 .setColumnFamily(ColumnFamilyDescriptorBuilder.of(cf1)).build(); 271 TableDescriptor t2 = TableDescriptorBuilder.newBuilder(TableName.valueOf("t2")) 272 .setColumnFamily(ColumnFamilyDescriptorBuilder.of(cf1)).build(); 273 RegionInfo hri1 = RegionInfoBuilder.newBuilder(t1.getTableName()).build(); 274 RegionInfo hri2 = RegionInfoBuilder.newBuilder(t2.getTableName()).build(); 275 276 List<ColumnFamilyDescriptor> cfs = new ArrayList<>(); 277 cfs.add(ColumnFamilyDescriptorBuilder.of(cf1)); 278 cfs.add(ColumnFamilyDescriptorBuilder.of(cf2)); 279 TableDescriptor t3 = 280 TableDescriptorBuilder.newBuilder(TableName.valueOf("t3")).setColumnFamilies(cfs).build(); 281 RegionInfo hri3 = RegionInfoBuilder.newBuilder(t3.getTableName()).build(); 282 283 // add edits and roll the wal 284 MultiVersionConcurrencyControl mvcc = new MultiVersionConcurrencyControl(); 285 NavigableMap<byte[], Integer> scopes1 = new TreeMap<>(Bytes.BYTES_COMPARATOR); 286 for (byte[] fam : t1.getColumnFamilyNames()) { 287 scopes1.put(fam, 0); 288 } 289 NavigableMap<byte[], Integer> scopes2 = new TreeMap<>(Bytes.BYTES_COMPARATOR); 290 for (byte[] fam : t2.getColumnFamilyNames()) { 291 scopes2.put(fam, 0); 292 } 293 NavigableMap<byte[], Integer> scopes3 = new TreeMap<>(Bytes.BYTES_COMPARATOR); 294 for (byte[] fam : t3.getColumnFamilyNames()) { 295 scopes3.put(fam, 0); 296 } 297 addEdits(wal, hri1, t1, 2, mvcc, scopes1, cf1); 298 wal.rollWriter(); 299 // add some more edits and roll the wal. This would reach the log number threshold 300 addEdits(wal, hri1, t1, 2, mvcc, scopes1, cf1); 301 wal.rollWriter(); 302 // with above rollWriter call, the max logs limit is reached. 303 waitNumRolledLogFiles(wal, 2); 304 305 // get the regions to flush; since there is only one region in the oldest wal, it should 306 // return only one region. 307 Map<byte[], List<byte[]>> regionsToFlush = wal.findRegionsToForceFlush(); 308 assertEquals(1, regionsToFlush.size()); 309 assertEquals(hri1.getEncodedNameAsBytes(), (byte[]) regionsToFlush.keySet().toArray()[0]); 310 // insert edits in second region 311 addEdits(wal, hri2, t2, 2, mvcc, scopes2, cf1); 312 // get the regions to flush, it should still read region1. 313 regionsToFlush = wal.findRegionsToForceFlush(); 314 assertEquals(1, regionsToFlush.size()); 315 assertEquals(hri1.getEncodedNameAsBytes(), (byte[]) regionsToFlush.keySet().toArray()[0]); 316 // flush region 1, and roll the wal file. Only last wal which has entries for region1 should 317 // remain. 318 flushRegion(wal, hri1.getEncodedNameAsBytes(), t1.getColumnFamilyNames()); 319 wal.rollWriter(); 320 // only one wal should remain now (that is for the second region). 321 waitNumRolledLogFiles(wal, 1); 322 // flush the second region 323 flushRegion(wal, hri2.getEncodedNameAsBytes(), t2.getColumnFamilyNames()); 324 wal.rollWriter(true); 325 // no wal should remain now. 326 waitNumRolledLogFiles(wal, 0); 327 // add edits both to region 1 and region 2, and roll. 328 addEdits(wal, hri1, t1, 2, mvcc, scopes1, cf1); 329 addEdits(wal, hri2, t2, 2, mvcc, scopes2, cf1); 330 wal.rollWriter(); 331 // add edits and roll the writer, to reach the max logs limit. 332 waitNumRolledLogFiles(wal, 1); 333 addEdits(wal, hri1, t1, 2, mvcc, scopes1, cf1); 334 wal.rollWriter(); 335 // it should return two regions to flush, as the oldest wal file has entries 336 // for both regions. 337 regionsToFlush = wal.findRegionsToForceFlush(); 338 assertEquals(2, regionsToFlush.size()); 339 // flush both regions 340 flushRegion(wal, hri1.getEncodedNameAsBytes(), t1.getColumnFamilyNames()); 341 flushRegion(wal, hri2.getEncodedNameAsBytes(), t2.getColumnFamilyNames()); 342 wal.rollWriter(true); 343 waitNumRolledLogFiles(wal, 0); 344 // Add an edit to region1, and roll the wal. 345 addEdits(wal, hri1, t1, 2, mvcc, scopes1, cf1); 346 // tests partial flush: roll on a partial flush, and ensure that wal is not archived. 347 wal.startCacheFlush(hri1.getEncodedNameAsBytes(), t1.getColumnFamilyNames()); 348 wal.rollWriter(); 349 wal.completeCacheFlush(hri1.getEncodedNameAsBytes(), HConstants.NO_SEQNUM); 350 waitNumRolledLogFiles(wal, 1); 351 352 // clear test data 353 flushRegion(wal, hri1.getEncodedNameAsBytes(), t1.getColumnFamilyNames()); 354 wal.rollWriter(true); 355 // add edits for three familes 356 addEdits(wal, hri3, t3, 2, mvcc, scopes3, cf1); 357 addEdits(wal, hri3, t3, 2, mvcc, scopes3, cf2); 358 addEdits(wal, hri3, t3, 2, mvcc, scopes3, cf3); 359 wal.rollWriter(); 360 addEdits(wal, hri3, t3, 2, mvcc, scopes3, cf1); 361 wal.rollWriter(); 362 waitNumRolledLogFiles(wal, 2); 363 // flush one family before archive oldest wal 364 Set<byte[]> flushedFamilyNames = new HashSet<>(); 365 flushedFamilyNames.add(Bytes.toBytes(cf1)); 366 flushRegion(wal, hri3.getEncodedNameAsBytes(), flushedFamilyNames); 367 regionsToFlush = wal.findRegionsToForceFlush(); 368 // then only two family need to be flushed when archive oldest wal 369 assertEquals(1, regionsToFlush.size()); 370 assertEquals(hri3.getEncodedNameAsBytes(), (byte[]) regionsToFlush.keySet().toArray()[0]); 371 assertEquals(2, regionsToFlush.get(hri3.getEncodedNameAsBytes()).size()); 372 } 373 374 /** 375 * On rolling a wal after reaching the threshold, {@link WAL#rollWriter()} returns the list of 376 * regions which should be flushed in order to archive the oldest wal file. 377 * <p> 378 * This method tests this behavior by inserting edits and rolling the wal enough times to reach 379 * the max number of logs threshold. It checks whether we get the "right regions and stores" for 380 * flush on rolling the wal. 381 */ 382 @Test 383 public void testFindMemStoresEligibleForFlush() throws Exception { 384 LOG.debug("testFindMemStoresEligibleForFlush"); 385 Configuration conf1 = HBaseConfiguration.create(CONF); 386 conf1.setInt("hbase.regionserver.maxlogs", 1); 387 try (AbstractFSWAL<?> wal = newWAL(FS, CommonFSUtils.getWALRootDir(conf1), DIR.toString(), 388 HConstants.HREGION_OLDLOGDIR_NAME, conf1, null, true, null, null)) { 389 testFindMemStoresEligibleForFlush(wal); 390 } 391 392 } 393 394 @Test(expected = IOException.class) 395 public void testFailedToCreateWALIfParentRenamed() 396 throws IOException, CommonFSUtils.StreamLacksCapabilityException { 397 AbstractFSWAL<?> wal = newWAL(FS, CommonFSUtils.getWALRootDir(CONF), 398 currentTest.getMethodName(), HConstants.HREGION_OLDLOGDIR_NAME, CONF, null, true, null, null); 399 long filenum = EnvironmentEdgeManager.currentTime(); 400 Path path = wal.computeFilename(filenum); 401 wal.createWriterInstance(FS, path); 402 Path parent = path.getParent(); 403 path = wal.computeFilename(filenum + 1); 404 Path newPath = new Path(parent.getParent(), parent.getName() + "-splitting"); 405 FS.rename(parent, newPath); 406 wal.createWriterInstance(FS, path); 407 fail("It should fail to create the new WAL"); 408 } 409 410 /** 411 * Test flush for sure has a sequence id that is beyond the last edit appended. We do this by 412 * slowing appends in the background ring buffer thread while in foreground we call flush. The 413 * addition of the sync over HRegion in flush should fix an issue where flush was returning before 414 * all of its appends had made it out to the WAL (HBASE-11109). 415 * @see <a href="https://issues.apache.org/jira/browse/HBASE-11109">HBASE-11109</a> 416 */ 417 @Test 418 public void testFlushSequenceIdIsGreaterThanAllEditsInHFile() throws IOException { 419 String testName = currentTest.getMethodName(); 420 final TableName tableName = TableName.valueOf(testName); 421 final RegionInfo hri = RegionInfoBuilder.newBuilder(tableName).build(); 422 final byte[] rowName = tableName.getName(); 423 final TableDescriptor htd = TableDescriptorBuilder.newBuilder(tableName) 424 .setColumnFamily(ColumnFamilyDescriptorBuilder.of("f")).build(); 425 HRegion r = HBaseTestingUtil.createRegionAndWAL(hri, TEST_UTIL.getDefaultRootDirPath(), 426 TEST_UTIL.getConfiguration(), htd); 427 HBaseTestingUtil.closeRegionAndWAL(r); 428 final int countPerFamily = 10; 429 final AtomicBoolean goslow = new AtomicBoolean(false); 430 NavigableMap<byte[], Integer> scopes = new TreeMap<>(Bytes.BYTES_COMPARATOR); 431 for (byte[] fam : htd.getColumnFamilyNames()) { 432 scopes.put(fam, 0); 433 } 434 // subclass and doctor a method. 435 AbstractFSWAL<?> wal = newSlowWAL(FS, CommonFSUtils.getWALRootDir(CONF), DIR.toString(), 436 testName, CONF, null, true, null, null, new Runnable() { 437 438 @Override 439 public void run() { 440 if (goslow.get()) { 441 Threads.sleep(100); 442 LOG.debug("Sleeping before appending 100ms"); 443 } 444 } 445 }); 446 HRegion region = HRegion.openHRegion(TEST_UTIL.getConfiguration(), 447 TEST_UTIL.getTestFileSystem(), TEST_UTIL.getDefaultRootDirPath(), hri, htd, wal); 448 EnvironmentEdge ee = EnvironmentEdgeManager.getDelegate(); 449 try { 450 List<Put> puts = null; 451 for (byte[] fam : htd.getColumnFamilyNames()) { 452 puts = TestWALReplay.addRegionEdits(rowName, fam, countPerFamily, ee, region, "x"); 453 } 454 455 // Now assert edits made it in. 456 final Get g = new Get(rowName); 457 Result result = region.get(g); 458 assertEquals(countPerFamily * htd.getColumnFamilyNames().size(), result.size()); 459 460 // Construct a WALEdit and add it a few times to the WAL. 461 WALEdit edits = new WALEdit(); 462 for (Put p : puts) { 463 ExtendedCellScanner cs = p.cellScanner(); 464 while (cs.advance()) { 465 WALEditInternalHelper.addExtendedCell(edits, cs.current()); 466 } 467 } 468 // Add any old cluster id. 469 List<UUID> clusterIds = new ArrayList<>(1); 470 clusterIds.add(TEST_UTIL.getRandomUUID()); 471 // Now make appends run slow. 472 goslow.set(true); 473 for (int i = 0; i < countPerFamily; i++) { 474 final RegionInfo info = region.getRegionInfo(); 475 final WALKeyImpl logkey = new WALKeyImpl(info.getEncodedNameAsBytes(), tableName, 476 EnvironmentEdgeManager.currentTime(), clusterIds, -1, -1, region.getMVCC(), scopes); 477 wal.append(info, logkey, edits, true); 478 region.getMVCC().completeAndWait(logkey.getWriteEntry()); 479 } 480 region.flush(true); 481 // FlushResult.flushSequenceId is not visible here so go get the current sequence id. 482 long currentSequenceId = region.getReadPoint(null); 483 // Now release the appends 484 goslow.set(false); 485 assertTrue(currentSequenceId >= region.getReadPoint(null)); 486 } finally { 487 region.close(true); 488 wal.close(); 489 } 490 } 491 492 @Test 493 public void testSyncNoAppend() throws IOException { 494 String testName = currentTest.getMethodName(); 495 AbstractFSWAL<?> wal = newWAL(FS, CommonFSUtils.getWALRootDir(CONF), DIR.toString(), testName, 496 CONF, null, true, null, null); 497 try { 498 wal.sync(); 499 } finally { 500 wal.close(); 501 } 502 } 503 504 @Test 505 public void testWriteEntryCanBeNull() throws IOException { 506 String testName = currentTest.getMethodName(); 507 AbstractFSWAL<?> wal = newWAL(FS, CommonFSUtils.getWALRootDir(CONF), DIR.toString(), testName, 508 CONF, null, true, null, null); 509 wal.close(); 510 TableDescriptor td = TableDescriptorBuilder.newBuilder(TableName.valueOf("table")) 511 .setColumnFamily(ColumnFamilyDescriptorBuilder.of("row")).build(); 512 RegionInfo ri = RegionInfoBuilder.newBuilder(td.getTableName()).build(); 513 MultiVersionConcurrencyControl mvcc = new MultiVersionConcurrencyControl(); 514 NavigableMap<byte[], Integer> scopes = new TreeMap<>(Bytes.BYTES_COMPARATOR); 515 for (byte[] fam : td.getColumnFamilyNames()) { 516 scopes.put(fam, 0); 517 } 518 long timestamp = EnvironmentEdgeManager.currentTime(); 519 byte[] row = Bytes.toBytes("row"); 520 WALEdit cols = new WALEdit(); 521 WALEditInternalHelper.addExtendedCell(cols, new KeyValue(row, row, row, timestamp, row)); 522 WALKeyImpl key = 523 new WALKeyImpl(ri.getEncodedNameAsBytes(), td.getTableName(), SequenceId.NO_SEQUENCE_ID, 524 timestamp, WALKey.EMPTY_UUIDS, HConstants.NO_NONCE, HConstants.NO_NONCE, mvcc, scopes); 525 try { 526 wal.append(ri, key, cols, true); 527 fail("Should fail since the wal has already been closed"); 528 } catch (IOException e) { 529 // expected 530 assertThat(e.getMessage(), containsString("log is closed")); 531 // the WriteEntry should be null since we fail before setting it. 532 assertNull(key.getWriteEntry()); 533 } 534 } 535 536 @Test(expected = WALClosedException.class) 537 public void testRollWriterForClosedWAL() throws IOException { 538 String testName = currentTest.getMethodName(); 539 AbstractFSWAL<?> wal = newWAL(FS, CommonFSUtils.getWALRootDir(CONF), DIR.toString(), testName, 540 CONF, null, true, null, null); 541 wal.close(); 542 wal.rollWriter(); 543 } 544 545 private AbstractFSWAL<?> createHoldingWAL(String testName, AtomicBoolean startHoldingForAppend, 546 CountDownLatch holdAppend) throws IOException { 547 FS.mkdirs(new Path(CommonFSUtils.getRootDir(CONF), testName)); 548 AbstractFSWAL<?> wal = newWAL(FS, CommonFSUtils.getRootDir(CONF), testName, 549 HConstants.HREGION_OLDLOGDIR_NAME, CONF, null, true, null, null); 550 // newWAL has already called wal.init() 551 wal.registerWALActionsListener(new WALActionsListener() { 552 @Override 553 public void visitLogEntryBeforeWrite(RegionInfo info, WALKey logKey, WALEdit logEdit) { 554 if (startHoldingForAppend.get()) { 555 try { 556 holdAppend.await(); 557 } catch (InterruptedException e) { 558 LOG.error(e.toString(), e); 559 } 560 } 561 } 562 }); 563 return wal; 564 } 565 566 private HRegion createHoldingHRegion(Configuration conf, TableDescriptor htd, WAL wal) 567 throws IOException { 568 RegionInfo hri = RegionInfoBuilder.newBuilder(htd.getTableName()).build(); 569 ChunkCreator.initialize(MemStoreLAB.CHUNK_SIZE_DEFAULT, false, 0, 0, 0, null, 570 MemStoreLAB.INDEX_CHUNK_SIZE_PERCENTAGE_DEFAULT); 571 TEST_UTIL.createLocalHRegion(hri, CONF, htd, wal).close(); 572 RegionServerServices rsServices = mock(RegionServerServices.class); 573 when(rsServices.getServerName()).thenReturn(ServerName.valueOf("localhost:12345", 123456)); 574 when(rsServices.getConfiguration()).thenReturn(conf); 575 return HRegion.openHRegion(TEST_UTIL.getDataTestDir(), hri, htd, wal, conf, rsServices, null); 576 } 577 578 private void doPutWithAsyncWAL(ExecutorService exec, HRegion region, Put put, 579 Runnable flushOrCloseRegion, AtomicBoolean startHoldingForAppend, 580 CountDownLatch flushOrCloseFinished, CountDownLatch holdAppend) 581 throws InterruptedException, IOException { 582 // do a regular write first because of memstore size calculation. 583 region.put(put); 584 585 startHoldingForAppend.set(true); 586 region.put(new Put(put).setDurability(Durability.ASYNC_WAL)); 587 588 // give the put a chance to start 589 Threads.sleep(3000); 590 591 exec.submit(flushOrCloseRegion); 592 593 // give the flush a chance to start. Flush should have got the region lock, and 594 // should have been waiting on the mvcc complete after this. 595 Threads.sleep(3000); 596 597 // let the append to WAL go through now that the flush already started 598 holdAppend.countDown(); 599 flushOrCloseFinished.await(); 600 } 601 602 // Testcase for HBASE-23181 603 @Test 604 public void testUnflushedSeqIdTrackingWithAsyncWal() throws IOException, InterruptedException { 605 String testName = currentTest.getMethodName(); 606 byte[] b = Bytes.toBytes("b"); 607 TableDescriptor htd = TableDescriptorBuilder.newBuilder(TableName.valueOf("table")) 608 .setColumnFamily(ColumnFamilyDescriptorBuilder.of(b)).build(); 609 610 AtomicBoolean startHoldingForAppend = new AtomicBoolean(false); 611 CountDownLatch holdAppend = new CountDownLatch(1); 612 CountDownLatch closeFinished = new CountDownLatch(1); 613 ExecutorService exec = Executors.newFixedThreadPool(1); 614 AbstractFSWAL<?> wal = createHoldingWAL(testName, startHoldingForAppend, holdAppend); 615 // open a new region which uses this WAL 616 HRegion region = createHoldingHRegion(TEST_UTIL.getConfiguration(), htd, wal); 617 try { 618 doPutWithAsyncWAL(exec, region, new Put(b).addColumn(b, b, b), () -> { 619 try { 620 Map<?, ?> closeResult = region.close(); 621 LOG.info("Close result:" + closeResult); 622 closeFinished.countDown(); 623 } catch (IOException e) { 624 LOG.error(e.toString(), e); 625 } 626 }, startHoldingForAppend, closeFinished, holdAppend); 627 628 // now check the region's unflushed seqIds. 629 long seqId = getEarliestMemStoreSeqNum(wal, region.getRegionInfo().getEncodedNameAsBytes()); 630 assertEquals("Found seqId for the region which is already closed", HConstants.NO_SEQNUM, 631 seqId); 632 } finally { 633 exec.shutdownNow(); 634 region.close(); 635 wal.close(); 636 } 637 } 638 639 public static long getEarliestMemStoreSeqNum(WAL wal, byte[] encodedRegionName) { 640 if (wal != null) { 641 if (wal instanceof AbstractFSWAL) { 642 return ((AbstractFSWAL<?>) wal).getSequenceIdAccounting() 643 .getLowestSequenceId(encodedRegionName); 644 } 645 } 646 return HConstants.NO_SEQNUM; 647 } 648 649 private static final Set<byte[]> STORES_TO_FLUSH = 650 Collections.newSetFromMap(new ConcurrentSkipListMap<byte[], Boolean>(Bytes.BYTES_COMPARATOR)); 651 652 // Testcase for HBASE-23157 653 @Test 654 public void testMaxFlushedSequenceIdGoBackwards() throws IOException, InterruptedException { 655 String testName = currentTest.getMethodName(); 656 byte[] a = Bytes.toBytes("a"); 657 byte[] b = Bytes.toBytes("b"); 658 TableDescriptor htd = TableDescriptorBuilder.newBuilder(TableName.valueOf("table")) 659 .setColumnFamily(ColumnFamilyDescriptorBuilder.of(a)) 660 .setColumnFamily(ColumnFamilyDescriptorBuilder.of(b)).build(); 661 662 AtomicBoolean startHoldingForAppend = new AtomicBoolean(false); 663 CountDownLatch holdAppend = new CountDownLatch(1); 664 CountDownLatch flushFinished = new CountDownLatch(1); 665 ExecutorService exec = Executors.newFixedThreadPool(2); 666 Configuration conf = new Configuration(TEST_UTIL.getConfiguration()); 667 conf.setClass(FlushPolicyFactory.HBASE_FLUSH_POLICY_KEY, FlushSpecificStoresPolicy.class, 668 FlushPolicy.class); 669 AbstractFSWAL<?> wal = createHoldingWAL(testName, startHoldingForAppend, holdAppend); 670 // open a new region which uses this WAL 671 HRegion region = createHoldingHRegion(conf, htd, wal); 672 try { 673 Put put = new Put(a).addColumn(a, a, a).addColumn(b, b, b); 674 doPutWithAsyncWAL(exec, region, put, () -> { 675 try { 676 HRegion.FlushResult flushResult = region.flush(true); 677 LOG.info("Flush result:" + flushResult.getResult()); 678 LOG.info("Flush succeeded:" + flushResult.isFlushSucceeded()); 679 flushFinished.countDown(); 680 } catch (IOException e) { 681 LOG.error(e.toString(), e); 682 } 683 }, startHoldingForAppend, flushFinished, holdAppend); 684 685 // get the max flushed sequence id after the first flush 686 long maxFlushedSeqId1 = region.getMaxFlushedSeqId(); 687 688 region.put(put); 689 // this time we only flush family a 690 STORES_TO_FLUSH.add(a); 691 region.flush(false); 692 693 // get the max flushed sequence id after the second flush 694 long maxFlushedSeqId2 = region.getMaxFlushedSeqId(); 695 // make sure that the maxFlushedSequenceId does not go backwards 696 assertTrue( 697 "maxFlushedSeqId1(" + maxFlushedSeqId1 698 + ") is not greater than or equal to maxFlushedSeqId2(" + maxFlushedSeqId2 + ")", 699 maxFlushedSeqId1 <= maxFlushedSeqId2); 700 } finally { 701 exec.shutdownNow(); 702 region.close(); 703 wal.close(); 704 } 705 } 706 707 public static final class FlushSpecificStoresPolicy extends FlushPolicy { 708 709 @Override 710 public Collection<HStore> selectStoresToFlush() { 711 if (STORES_TO_FLUSH.isEmpty()) { 712 return region.getStores(); 713 } else { 714 return STORES_TO_FLUSH.stream().map(region::getStore).collect(Collectors.toList()); 715 } 716 } 717 } 718}