View Javadoc

1   /**
2    * Copyright 2010 The Apache Software Foundation
3    *
4    * Licensed to the Apache Software Foundation (ASF) under one
5    * or more contributor license agreements.  See the NOTICE file
6    * distributed with this work for additional information
7    * regarding copyright ownership.  The ASF licenses this file
8    * to you under the Apache License, Version 2.0 (the
9    * "License"); you may not use this file except in compliance
10   * with the License.  You may obtain a copy of the License at
11   *
12   *     http://www.apache.org/licenses/LICENSE-2.0
13   *
14   * Unless required by applicable law or agreed to in writing, software
15   * distributed under the License is distributed on an "AS IS" BASIS,
16   * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
17   * See the License for the specific language governing permissions and
18   * limitations under the License.
19   */
20  
21  package org.apache.hadoop.hbase.regionserver;
22  
23  import java.io.IOException;
24  import java.util.NavigableSet;
25  
26  import org.apache.hadoop.hbase.HConstants;
27  import org.apache.hadoop.hbase.KeyValue;
28  import org.apache.hadoop.hbase.client.Scan;
29  import org.apache.hadoop.hbase.filter.Filter;
30  import org.apache.hadoop.hbase.filter.Filter.ReturnCode;
31  import org.apache.hadoop.hbase.io.TimeRange;
32  import org.apache.hadoop.hbase.regionserver.DeleteTracker.DeleteResult;
33  import org.apache.hadoop.hbase.util.Bytes;
34  import org.apache.hadoop.hbase.util.EnvironmentEdgeManager;
35  
36  /**
37   * A query matcher that is specifically designed for the scan case.
38   */
39  public class ScanQueryMatcher {
40    // Optimization so we can skip lots of compares when we decide to skip
41    // to the next row.
42    private boolean stickyNextRow;
43    private final byte[] stopRow;
44  
45    private final TimeRange tr;
46  
47    private final Filter filter;
48  
49    /** Keeps track of deletes */
50    private final DeleteTracker deletes;
51  
52    /*
53     * The following three booleans define how we deal with deletes.
54     * There are three different aspects:
55     * 1. Whether to keep delete markers. This is used in compactions.
56     *    Minor compactions always keep delete markers.
57     * 2. Whether to keep deleted rows. This is also used in compactions,
58     *    if the store is set to keep deleted rows. This implies keeping
59     *    the delete markers as well.
60     *    In this case deleted rows are subject to the normal max version
61     *    and TTL/min version rules just like "normal" rows.
62     * 3. Whether a scan can do time travel queries even before deleted
63     *    marker to reach deleted rows.
64     */
65    /** whether to retain delete markers */
66    private final boolean retainDeletesInOutput;
67    /** whether to return deleted rows */
68    private final boolean keepDeletedCells;
69    /** whether time range queries can see rows "behind" a delete */
70    private final boolean seePastDeleteMarkers;
71  
72  
73    /** Keeps track of columns and versions */
74    private final ColumnTracker columns;
75  
76    /** Key to seek to in memstore and StoreFiles */
77    private final KeyValue startKey;
78  
79    /** Row comparator for the region this query is for */
80    private final KeyValue.KeyComparator rowComparator;
81  
82    /* row is not private for tests */
83    /** Row the query is on */
84    byte [] row;
85    int rowOffset;
86    short rowLength;
87    
88    /**
89     * Oldest put in any of the involved store files
90     * Used to decide whether it is ok to delete
91     * family delete marker of this store keeps
92     * deleted KVs.
93     */
94    private final long earliestPutTs;
95  
96    /** readPoint over which the KVs are unconditionally included */
97    protected long maxReadPointToTrackVersions;
98  
99    /**
100    * This variable shows whether there is an null column in the query. There
101    * always exists a null column in the wildcard column query.
102    * There maybe exists a null column in the explicit column query based on the
103    * first column.
104    * */
105   private boolean hasNullColumn = true;
106 
107   // By default, when hbase.hstore.time.to.purge.deletes is 0ms, a delete
108   // marker is always removed during a major compaction. If set to non-zero
109   // value then major compaction will try to keep a delete marker around for
110   // the given number of milliseconds. We want to keep the delete markers
111   // around a bit longer because old puts might appear out-of-order. For
112   // example, during log replication between two clusters.
113   //
114   // If the delete marker has lived longer than its column-family's TTL then
115   // the delete marker will be removed even if time.to.purge.deletes has not
116   // passed. This is because all the Puts that this delete marker can influence
117   // would have also expired. (Removing of delete markers on col family TTL will
118   // not happen if min-versions is set to non-zero)
119   //
120   // But, if time.to.purge.deletes has not expired then a delete
121   // marker will not be removed just because there are no Puts that it is
122   // currently influencing. This is because Puts, that this delete can
123   // influence.  may appear out of order.
124   private final long timeToPurgeDeletes;
125   
126   private final boolean isUserScan;
127 
128   /**
129    * Construct a QueryMatcher for a scan
130    * @param scan
131    * @param scanInfo The store's immutable scan info
132    * @param columns
133    * @param scanType Type of the scan
134    * @param earliestPutTs Earliest put seen in any of the store files.
135    * @param oldestUnexpiredTS the oldest timestamp we are interested in,
136    *  based on TTL
137    */
138   public ScanQueryMatcher(Scan scan, Store.ScanInfo scanInfo,
139       NavigableSet<byte[]> columns, ScanType scanType,
140       long readPointToUse, long earliestPutTs, long oldestUnexpiredTS) {
141     this.tr = scan.getTimeRange();
142     this.rowComparator = scanInfo.getComparator().getRawComparator();
143     this.deletes =  new ScanDeleteTracker();
144     this.stopRow = scan.getStopRow();
145     this.startKey = KeyValue.createFirstDeleteFamilyOnRow(scan.getStartRow(),
146         scanInfo.getFamily());
147     this.filter = scan.getFilter();
148     this.earliestPutTs = earliestPutTs;
149     this.maxReadPointToTrackVersions = readPointToUse;
150     this.timeToPurgeDeletes = scanInfo.getTimeToPurgeDeletes();
151 
152     /* how to deal with deletes */
153     this.isUserScan = scanType == ScanType.USER_SCAN;
154     // keep deleted cells: if compaction or raw scan
155     this.keepDeletedCells = (scanInfo.getKeepDeletedCells() && !isUserScan) || scan.isRaw();
156     // retain deletes: if minor compaction or raw scan
157     this.retainDeletesInOutput = scanType == ScanType.MINOR_COMPACT || scan.isRaw();
158     // seePastDeleteMarker: user initiated scans
159     this.seePastDeleteMarkers = scanInfo.getKeepDeletedCells() && isUserScan;
160 
161     int maxVersions =
162         scan.isRaw() ? scan.getMaxVersions() : Math.min(scan.getMaxVersions(),
163           scanInfo.getMaxVersions());
164 
165     // Single branch to deal with two types of reads (columns vs all in family)
166     if (columns == null || columns.size() == 0) {
167       // there is always a null column in the wildcard column query.
168       hasNullColumn = true;
169 
170       // use a specialized scan for wildcard column tracker.
171       this.columns = new ScanWildcardColumnTracker(
172           scanInfo.getMinVersions(), maxVersions, oldestUnexpiredTS);
173     } else {
174       // whether there is null column in the explicit column query
175       hasNullColumn = (columns.first().length == 0);
176 
177       // We can share the ExplicitColumnTracker, diff is we reset
178       // between rows, not between storefiles.
179       byte[] attr = scan.getAttribute(Scan.HINT_LOOKAHEAD);
180       this.columns = new ExplicitColumnTracker(columns, scanInfo.getMinVersions(), maxVersions,
181           oldestUnexpiredTS, attr == null ? 0 : Bytes.toInt(attr));
182     }
183   }
184 
185   /*
186    * Constructor for tests
187    */
188   ScanQueryMatcher(Scan scan, Store.ScanInfo scanInfo,
189       NavigableSet<byte[]> columns, long oldestUnexpiredTS) {
190     this(scan, scanInfo, columns, ScanType.USER_SCAN,
191           Long.MAX_VALUE, /* max Readpoint to track versions */
192         HConstants.LATEST_TIMESTAMP, oldestUnexpiredTS);
193   }
194 
195   /**
196    *
197    * @return  whether there is an null column in the query
198    */
199   public boolean hasNullColumnInQuery() {
200     return hasNullColumn;
201   }
202 
203   /**
204    * Determines if the caller should do one of several things:
205    * - seek/skip to the next row (MatchCode.SEEK_NEXT_ROW)
206    * - seek/skip to the next column (MatchCode.SEEK_NEXT_COL)
207    * - include the current KeyValue (MatchCode.INCLUDE)
208    * - ignore the current KeyValue (MatchCode.SKIP)
209    * - got to the next row (MatchCode.DONE)
210    *
211    * @param kv KeyValue to check
212    * @return The match code instance.
213    * @throws IOException in case there is an internal consistency problem
214    *      caused by a data corruption.
215    */
216   public MatchCode match(KeyValue kv) throws IOException {
217     if (filter != null && filter.filterAllRemaining()) {
218       return MatchCode.DONE_SCAN;
219     }
220 
221     byte [] bytes = kv.getBuffer();
222     int offset = kv.getOffset();
223 
224     int keyLength = Bytes.toInt(bytes, offset, Bytes.SIZEOF_INT);
225     offset += KeyValue.ROW_OFFSET;
226 
227     int initialOffset = offset;
228 
229     short rowLength = Bytes.toShort(bytes, offset, Bytes.SIZEOF_SHORT);
230     offset += Bytes.SIZEOF_SHORT;
231 
232     int ret = this.rowComparator.compareRows(row, this.rowOffset, this.rowLength,
233         bytes, offset, rowLength);
234     if (ret <= -1) {
235       return MatchCode.DONE;
236     } else if (ret >= 1) {
237       // could optimize this, if necessary?
238       // Could also be called SEEK_TO_CURRENT_ROW, but this
239       // should be rare/never happens.
240       return MatchCode.SEEK_NEXT_ROW;
241     }
242 
243     // optimize case.
244     if (this.stickyNextRow)
245         return MatchCode.SEEK_NEXT_ROW;
246 
247     if (this.columns.done()) {
248       stickyNextRow = true;
249       return MatchCode.SEEK_NEXT_ROW;
250     }
251 
252     //Passing rowLength
253     offset += rowLength;
254 
255     //Skipping family
256     byte familyLength = bytes [offset];
257     offset += familyLength + 1;
258 
259     int qualLength = keyLength -
260       (offset - initialOffset) - KeyValue.TIMESTAMP_TYPE_SIZE;
261 
262     long timestamp = Bytes.toLong(bytes, initialOffset + keyLength - KeyValue.TIMESTAMP_TYPE_SIZE);
263     // check for early out based on timestamp alone
264     if (columns.isDone(timestamp)) {
265         return columns.getNextRowOrNextColumn(bytes, offset, qualLength);
266     }
267 
268     /*
269      * The delete logic is pretty complicated now.
270      * This is corroborated by the following:
271      * 1. The store might be instructed to keep deleted rows around.
272      * 2. A scan can optionally see past a delete marker now.
273      * 3. If deleted rows are kept, we have to find out when we can
274      *    remove the delete markers.
275      * 4. Family delete markers are always first (regardless of their TS)
276      * 5. Delete markers should not be counted as version
277      * 6. Delete markers affect puts of the *same* TS
278      * 7. Delete marker need to be version counted together with puts
279      *    they affect
280      */
281     byte type = bytes[initialOffset + keyLength - 1];
282     if (kv.isDelete()) {
283       if (!keepDeletedCells) {
284         // first ignore delete markers if the scanner can do so, and the
285         // range does not include the marker
286         //
287         // during flushes and compactions also ignore delete markers newer
288         // than the readpoint of any open scanner, this prevents deleted
289         // rows that could still be seen by a scanner from being collected
290         boolean includeDeleteMarker = seePastDeleteMarkers ?
291             tr.withinTimeRange(timestamp) :
292             tr.withinOrAfterTimeRange(timestamp);
293         if (includeDeleteMarker
294             && kv.getMemstoreTS() <= maxReadPointToTrackVersions) {
295           this.deletes.add(bytes, offset, qualLength, timestamp, type);
296         }
297         // Can't early out now, because DelFam come before any other keys
298       }
299       if (retainDeletesInOutput
300           || (!isUserScan && (EnvironmentEdgeManager.currentTimeMillis() - timestamp) <= timeToPurgeDeletes)
301           || kv.getMemstoreTS() > maxReadPointToTrackVersions) {
302         // always include or it is not time yet to check whether it is OK
303         // to purge deltes or not
304         if (!isUserScan) {
305           // if this is not a user scan (compaction), we can filter this deletemarker right here
306           // otherwise (i.e. a "raw" scan) we fall through to normal version and timerange checking
307           return MatchCode.INCLUDE;
308         }
309       } else if (keepDeletedCells) {
310         if (timestamp < earliestPutTs) {
311           // keeping delete rows, but there are no puts older than
312           // this delete in the store files.
313           return columns.getNextRowOrNextColumn(bytes, offset, qualLength);
314         }
315         // else: fall through and do version counting on the
316         // delete markers
317       } else {
318         return MatchCode.SKIP;
319       }
320       // note the following next else if...
321       // delete marker are not subject to other delete markers
322     } else if (!this.deletes.isEmpty()) {
323       DeleteResult deleteResult = deletes.isDeleted(bytes, offset, qualLength,
324           timestamp);
325       switch (deleteResult) {
326         case FAMILY_DELETED:
327         case COLUMN_DELETED:
328           return columns.getNextRowOrNextColumn(bytes, offset, qualLength);
329         case VERSION_DELETED:
330           return MatchCode.SKIP;
331         case NOT_DELETED:
332           break;
333         default:
334           throw new RuntimeException("UNEXPECTED");
335         }
336     }
337 
338     int timestampComparison = tr.compare(timestamp);
339     if (timestampComparison >= 1) {
340       return MatchCode.SKIP;
341     } else if (timestampComparison <= -1) {
342       return columns.getNextRowOrNextColumn(bytes, offset, qualLength);
343     }
344 
345     // STEP 1: Check if the column is part of the requested columns
346     MatchCode colChecker = columns.checkColumn(bytes, offset, qualLength, type);
347     if (colChecker == MatchCode.INCLUDE) {
348       ReturnCode filterResponse = ReturnCode.SKIP;
349       // STEP 2: Yes, the column is part of the requested columns. Check if filter is present
350       if (filter != null) {
351         // STEP 3: Filter the key value and return if it filters out
352         filterResponse = filter.filterKeyValue(kv);
353         switch (filterResponse) {
354         case SKIP:
355           return MatchCode.SKIP;
356         case NEXT_COL:
357           return columns.getNextRowOrNextColumn(bytes, offset, qualLength);
358         case NEXT_ROW:
359           stickyNextRow = true;
360           return MatchCode.SEEK_NEXT_ROW;
361         case SEEK_NEXT_USING_HINT:
362           return MatchCode.SEEK_NEXT_USING_HINT;
363         default:
364           //It means it is either include or include and seek next
365           break;
366         }
367       }
368       /*
369        * STEP 4: Reaching this step means the column is part of the requested columns and either
370        * the filter is null or the filter has returned INCLUDE or INCLUDE_AND_NEXT_COL response.
371        * Now check the number of versions needed. This method call returns SKIP, INCLUDE,
372        * INCLUDE_AND_SEEK_NEXT_ROW, INCLUDE_AND_SEEK_NEXT_COL.
373        *
374        * FilterResponse            ColumnChecker               Desired behavior
375        * INCLUDE                   SKIP                        row has already been included, SKIP.
376        * INCLUDE                   INCLUDE                     INCLUDE
377        * INCLUDE                   INCLUDE_AND_SEEK_NEXT_COL   INCLUDE_AND_SEEK_NEXT_COL
378        * INCLUDE                   INCLUDE_AND_SEEK_NEXT_ROW   INCLUDE_AND_SEEK_NEXT_ROW
379        * INCLUDE_AND_SEEK_NEXT_COL SKIP                        row has already been included, SKIP.
380        * INCLUDE_AND_SEEK_NEXT_COL INCLUDE                     INCLUDE_AND_SEEK_NEXT_COL
381        * INCLUDE_AND_SEEK_NEXT_COL INCLUDE_AND_SEEK_NEXT_COL   INCLUDE_AND_SEEK_NEXT_COL
382        * INCLUDE_AND_SEEK_NEXT_COL INCLUDE_AND_SEEK_NEXT_ROW   INCLUDE_AND_SEEK_NEXT_ROW
383        *
384        * In all the above scenarios, we return the column checker return value except for
385        * FilterResponse (INCLUDE_AND_SEEK_NEXT_COL) and ColumnChecker(INCLUDE)
386        */
387       colChecker =
388           columns.checkVersions(bytes, offset, qualLength, timestamp, type,
389             kv.getMemstoreTS() > maxReadPointToTrackVersions);
390       //Optimize with stickyNextRow
391       stickyNextRow = colChecker == MatchCode.INCLUDE_AND_SEEK_NEXT_ROW ? true : stickyNextRow;
392       return (filterResponse == ReturnCode.INCLUDE_AND_NEXT_COL &&
393           colChecker == MatchCode.INCLUDE) ? MatchCode.INCLUDE_AND_SEEK_NEXT_COL
394           : colChecker;
395     }
396     stickyNextRow = (colChecker == MatchCode.SEEK_NEXT_ROW) ? true
397         : stickyNextRow;
398     return colChecker;
399   }
400 
401   public boolean moreRowsMayExistAfter(KeyValue kv) {
402     if (!Bytes.equals(stopRow , HConstants.EMPTY_END_ROW) &&
403         rowComparator.compareRows(kv.getBuffer(),kv.getRowOffset(),
404             kv.getRowLength(), stopRow, 0, stopRow.length) >= 0) {
405       // KV >= STOPROW
406       // then NO there is nothing left.
407       return false;
408     } else {
409       return true;
410     }
411   }
412 
413   /**
414    * Set current row
415    * @param row
416    */
417   public void setRow(byte [] row, int offset, short length) {
418     this.row = row;
419     this.rowOffset = offset;
420     this.rowLength = length;
421     reset();
422   }
423 
424   public void reset() {
425     this.deletes.reset();
426     this.columns.reset();
427 
428     stickyNextRow = false;
429   }
430 
431   /**
432    *
433    * @return the start key
434    */
435   public KeyValue getStartKey() {
436     return this.startKey;
437   }
438 
439   /**
440    *
441    * @return the Filter
442    */
443   Filter getFilter() {
444     return this.filter;
445   }
446 
447   public KeyValue getNextKeyHint(KeyValue kv) {
448     if (filter == null) {
449       return null;
450     } else {
451       return filter.getNextKeyHint(kv);
452     }
453   }
454 
455   public KeyValue getKeyForNextColumn(KeyValue kv) {
456     ColumnCount nextColumn = columns.getColumnHint();
457     if (nextColumn == null) {
458       return KeyValue.createLastOnRow(
459           kv.getBuffer(), kv.getRowOffset(), kv.getRowLength(),
460           kv.getBuffer(), kv.getFamilyOffset(), kv.getFamilyLength(),
461           kv.getBuffer(), kv.getQualifierOffset(), kv.getQualifierLength());
462     } else {
463       return KeyValue.createFirstOnRow(
464           kv.getBuffer(), kv.getRowOffset(), kv.getRowLength(),
465           kv.getBuffer(), kv.getFamilyOffset(), kv.getFamilyLength(),
466           nextColumn.getBuffer(), nextColumn.getOffset(), nextColumn.getLength());
467     }
468   }
469 
470   public KeyValue getKeyForNextRow(KeyValue kv) {
471     return KeyValue.createLastOnRow(
472         kv.getBuffer(), kv.getRowOffset(), kv.getRowLength(),
473         null, 0, 0,
474         null, 0, 0);
475   }
476 
477   // Used only for testing purposes
478   static MatchCode checkColumn(ColumnTracker columnTracker, byte[] bytes, int offset, int length,
479       long ttl, byte type, boolean ignoreCount) throws IOException {
480     MatchCode matchCode = columnTracker.checkColumn(bytes, offset, length, type);
481     if (matchCode == MatchCode.INCLUDE) {
482       return columnTracker.checkVersions(bytes, offset, length, ttl, type, ignoreCount);
483     }
484     return matchCode;
485   }
486 
487   /**
488    * {@link #match} return codes.  These instruct the scanner moving through
489    * memstores and StoreFiles what to do with the current KeyValue.
490    * <p>
491    * Additionally, this contains "early-out" language to tell the scanner to
492    * move on to the next File (memstore or Storefile), or to return immediately.
493    */
494   public static enum MatchCode {
495     /**
496      * Include KeyValue in the returned result
497      */
498     INCLUDE,
499 
500     /**
501      * Do not include KeyValue in the returned result
502      */
503     SKIP,
504 
505     /**
506      * Do not include, jump to next StoreFile or memstore (in time order)
507      */
508     NEXT,
509 
510     /**
511      * Do not include, return current result
512      */
513     DONE,
514 
515     /**
516      * These codes are used by the ScanQueryMatcher
517      */
518 
519     /**
520      * Done with the row, seek there.
521      */
522     SEEK_NEXT_ROW,
523     /**
524      * Done with column, seek to next.
525      */
526     SEEK_NEXT_COL,
527 
528     /**
529      * Done with scan, thanks to the row filter.
530      */
531     DONE_SCAN,
532 
533     /*
534      * Seek to next key which is given as hint.
535      */
536     SEEK_NEXT_USING_HINT,
537 
538     /**
539      * Include KeyValue and done with column, seek to next.
540      */
541     INCLUDE_AND_SEEK_NEXT_COL,
542 
543     /**
544      * Include KeyValue and done with row, seek to next.
545      */
546     INCLUDE_AND_SEEK_NEXT_ROW,
547   }
548 }