1 /** 2 * Licensed to the Apache Software Foundation (ASF) under one 3 * or more contributor license agreements. See the NOTICE file 4 * distributed with this work for additional information 5 * regarding copyright ownership. The ASF licenses this file 6 * to you under the Apache License, Version 2.0 (the 7 * "License"); you may not use this file except in compliance 8 * with the License. You may obtain a copy of the License at 9 * 10 * http://www.apache.org/licenses/LICENSE-2.0 11 * 12 * Unless required by applicable law or agreed to in writing, software 13 * distributed under the License is distributed on an "AS IS" BASIS, 14 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 15 * See the License for the specific language governing permissions and 16 * limitations under the License. 17 */ 18 package org.apache.hadoop.hbase.regionserver; 19 20 import java.io.IOException; 21 import java.util.Collection; 22 import java.util.List; 23 import java.util.NavigableSet; 24 25 import org.apache.hadoop.classification.InterfaceAudience; 26 import org.apache.hadoop.classification.InterfaceStability; 27 import org.apache.hadoop.fs.FileSystem; 28 import org.apache.hadoop.fs.Path; 29 import org.apache.hadoop.hbase.Cell; 30 import org.apache.hadoop.hbase.TableName; 31 import org.apache.hadoop.hbase.HColumnDescriptor; 32 import org.apache.hadoop.hbase.HRegionInfo; 33 import org.apache.hadoop.hbase.KeyValue; 34 import org.apache.hadoop.hbase.client.Scan; 35 import org.apache.hadoop.hbase.io.HeapSize; 36 import org.apache.hadoop.hbase.io.compress.Compression; 37 import org.apache.hadoop.hbase.io.hfile.CacheConfig; 38 import org.apache.hadoop.hbase.io.hfile.HFileDataBlockEncoder; 39 import org.apache.hadoop.hbase.protobuf.generated.WALProtos.CompactionDescriptor; 40 import org.apache.hadoop.hbase.regionserver.compactions.CompactionContext; 41 import org.apache.hadoop.hbase.regionserver.compactions.CompactionProgress; 42 import org.apache.hadoop.hbase.regionserver.compactions.CompactionRequest; 43 44 /** 45 * Interface for objects that hold a column family in a Region. Its a memstore and a set of zero or 46 * more StoreFiles, which stretch backwards over time. 47 */ 48 @InterfaceAudience.Private 49 @InterfaceStability.Evolving 50 public interface Store extends HeapSize, StoreConfigInformation { 51 52 /* The default priority for user-specified compaction requests. 53 * The user gets top priority unless we have blocking compactions. (Pri <= 0) 54 */ int PRIORITY_USER = 1; 55 int NO_PRIORITY = Integer.MIN_VALUE; 56 57 // General Accessors 58 KeyValue.KVComparator getComparator(); 59 60 Collection<StoreFile> getStorefiles(); 61 62 /** 63 * Close all the readers We don't need to worry about subsequent requests because the HRegion 64 * holds a write lock that will prevent any more reads or writes. 65 * @return the {@link StoreFile StoreFiles} that were previously being used. 66 * @throws IOException on failure 67 */ 68 Collection<StoreFile> close() throws IOException; 69 70 /** 71 * Return a scanner for both the memstore and the HStore files. Assumes we are not in a 72 * compaction. 73 * @param scan Scan to apply when scanning the stores 74 * @param targetCols columns to scan 75 * @return a scanner over the current key values 76 * @throws IOException on failure 77 */ 78 KeyValueScanner getScanner(Scan scan, final NavigableSet<byte[]> targetCols) 79 throws IOException; 80 81 /** 82 * Get all scanners with no filtering based on TTL (that happens further down 83 * the line). 84 * @param cacheBlocks 85 * @param isGet 86 * @param isCompaction 87 * @param matcher 88 * @param startRow 89 * @param stopRow 90 * @return all scanners for this store 91 */ 92 List<KeyValueScanner> getScanners( 93 boolean cacheBlocks, 94 boolean isGet, 95 boolean isCompaction, 96 ScanQueryMatcher matcher, 97 byte[] startRow, 98 byte[] stopRow 99 ) throws IOException; 100 101 ScanInfo getScanInfo(); 102 103 /** 104 * Adds or replaces the specified KeyValues. 105 * <p> 106 * For each KeyValue specified, if a cell with the same row, family, and qualifier exists in 107 * MemStore, it will be replaced. Otherwise, it will just be inserted to MemStore. 108 * <p> 109 * This operation is atomic on each KeyValue (row/family/qualifier) but not necessarily atomic 110 * across all of them. 111 * @param cells 112 * @param readpoint readpoint below which we can safely remove duplicate KVs 113 * @return memstore size delta 114 * @throws IOException 115 */ 116 long upsert(Iterable<Cell> cells, long readpoint) throws IOException; 117 118 /** 119 * Adds a value to the memstore 120 * @param kv 121 * @return memstore size delta 122 */ 123 long add(KeyValue kv); 124 125 /** 126 * When was the last edit done in the memstore 127 */ 128 long timeOfOldestEdit(); 129 130 /** 131 * Removes a kv from the memstore. The KeyValue is removed only if its key & memstoreTS match the 132 * key & memstoreTS value of the kv parameter. 133 * @param kv 134 */ 135 void rollback(final KeyValue kv); 136 137 /** 138 * Find the key that matches <i>row</i> exactly, or the one that immediately precedes it. WARNING: 139 * Only use this method on a table where writes occur with strictly increasing timestamps. This 140 * method assumes this pattern of writes in order to make it reasonably performant. Also our 141 * search is dependent on the axiom that deletes are for cells that are in the container that 142 * follows whether a memstore snapshot or a storefile, not for the current container: i.e. we'll 143 * see deletes before we come across cells we are to delete. Presumption is that the 144 * memstore#kvset is processed before memstore#snapshot and so on. 145 * @param row The row key of the targeted row. 146 * @return Found keyvalue or null if none found. 147 * @throws IOException 148 */ 149 KeyValue getRowKeyAtOrBefore(final byte[] row) throws IOException; 150 151 FileSystem getFileSystem(); 152 153 /* 154 * @param maxKeyCount 155 * @param compression Compression algorithm to use 156 * @param isCompaction whether we are creating a new file in a compaction 157 * @param includeMVCCReadpoint whether we should out the MVCC readpoint 158 * @return Writer for a new StoreFile in the tmp dir. 159 */ 160 StoreFile.Writer createWriterInTmp( 161 long maxKeyCount, 162 Compression.Algorithm compression, 163 boolean isCompaction, 164 boolean includeMVCCReadpoint 165 ) throws IOException; 166 167 // Compaction oriented methods 168 169 boolean throttleCompaction(long compactionSize); 170 171 /** 172 * getter for CompactionProgress object 173 * @return CompactionProgress object; can be null 174 */ 175 CompactionProgress getCompactionProgress(); 176 177 CompactionContext requestCompaction() throws IOException; 178 179 CompactionContext requestCompaction(int priority, CompactionRequest baseRequest) 180 throws IOException; 181 182 void cancelRequestedCompaction(CompactionContext compaction); 183 184 List<StoreFile> compact(CompactionContext compaction) throws IOException; 185 186 /** 187 * @return true if we should run a major compaction. 188 */ 189 boolean isMajorCompaction() throws IOException; 190 191 void triggerMajorCompaction(); 192 193 /** 194 * See if there's too much store files in this store 195 * @return true if number of store files is greater than the number defined in minFilesToCompact 196 */ 197 boolean needsCompaction(); 198 199 int getCompactPriority(); 200 201 StoreFlushContext createFlushContext(long cacheFlushId); 202 203 /** 204 * Call to complete a compaction. Its for the case where we find in the WAL a compaction 205 * that was not finished. We could find one recovering a WAL after a regionserver crash. 206 * See HBASE-2331. 207 * @param compaction 208 */ 209 void completeCompactionMarker(CompactionDescriptor compaction) 210 throws IOException; 211 212 // Split oriented methods 213 214 boolean canSplit(); 215 216 /** 217 * Determines if Store should be split 218 * @return byte[] if store should be split, null otherwise. 219 */ 220 byte[] getSplitPoint(); 221 222 // Bulk Load methods 223 224 /** 225 * This throws a WrongRegionException if the HFile does not fit in this region, or an 226 * InvalidHFileException if the HFile is not valid. 227 */ 228 void assertBulkLoadHFileOk(Path srcPath) throws IOException; 229 230 /** 231 * This method should only be called from HRegion. It is assumed that the ranges of values in the 232 * HFile fit within the stores assigned region. (assertBulkLoadHFileOk checks this) 233 * 234 * @param srcPathStr 235 * @param sequenceId sequence Id associated with the HFile 236 */ 237 void bulkLoadHFile(String srcPathStr, long sequenceId) throws IOException; 238 239 // General accessors into the state of the store 240 // TODO abstract some of this out into a metrics class 241 242 /** 243 * @return <tt>true</tt> if the store has any underlying reference files to older HFiles 244 */ 245 boolean hasReferences(); 246 247 /** 248 * @return The size of this store's memstore, in bytes 249 */ 250 long getMemStoreSize(); 251 252 HColumnDescriptor getFamily(); 253 254 /** 255 * @return The maximum memstoreTS in all store files. 256 */ 257 long getMaxMemstoreTS(); 258 259 /** 260 * @return the data block encoder 261 */ 262 HFileDataBlockEncoder getDataBlockEncoder(); 263 264 /** @return aggregate size of all HStores used in the last compaction */ 265 long getLastCompactSize(); 266 267 /** @return aggregate size of HStore */ 268 long getSize(); 269 270 /** 271 * @return Count of store files 272 */ 273 int getStorefilesCount(); 274 275 /** 276 * @return The size of the store files, in bytes, uncompressed. 277 */ 278 long getStoreSizeUncompressed(); 279 280 /** 281 * @return The size of the store files, in bytes. 282 */ 283 long getStorefilesSize(); 284 285 /** 286 * @return The size of the store file indexes, in bytes. 287 */ 288 long getStorefilesIndexSize(); 289 290 /** 291 * Returns the total size of all index blocks in the data block indexes, including the root level, 292 * intermediate levels, and the leaf level for multi-level indexes, or just the root level for 293 * single-level indexes. 294 * @return the total size of block indexes in the store 295 */ 296 long getTotalStaticIndexSize(); 297 298 /** 299 * Returns the total byte size of all Bloom filter bit arrays. For compound Bloom filters even the 300 * Bloom blocks currently not loaded into the block cache are counted. 301 * @return the total size of all Bloom filters in the store 302 */ 303 long getTotalStaticBloomSize(); 304 305 // Test-helper methods 306 307 /** 308 * Used for tests. 309 * @return cache configuration for this Store. 310 */ 311 CacheConfig getCacheConfig(); 312 313 /** 314 * @return the parent region info hosting this store 315 */ 316 HRegionInfo getRegionInfo(); 317 318 RegionCoprocessorHost getCoprocessorHost(); 319 320 boolean areWritesEnabled(); 321 322 /** 323 * @return The smallest mvcc readPoint across all the scanners in this 324 * region. Writes older than this readPoint, are included in every 325 * read operation. 326 */ 327 long getSmallestReadPoint(); 328 329 String getColumnFamilyName(); 330 331 TableName getTableName(); 332 333 /* 334 * @param o Observer who wants to know about changes in set of Readers 335 */ 336 void addChangedReaderObserver(ChangedReadersObserver o); 337 338 /* 339 * @param o Observer no longer interested in changes in set of Readers. 340 */ 341 void deleteChangedReaderObserver(ChangedReadersObserver o); 342 343 /** 344 * @return Whether this store has too many store files. 345 */ 346 boolean hasTooManyStoreFiles(); 347 }