1
|
|
|
<?php |
2
|
|
|
declare(strict_types=1); |
3
|
|
|
|
4
|
|
|
namespace Level23\Druid\Context; |
5
|
|
|
|
6
|
|
|
/** |
7
|
|
|
* GroupBy queries can be executed using two different strategies. The default strategy for a cluster is determined |
8
|
|
|
* by the "druid.query.groupBy.defaultStrategy" runtime property on the Broker. This can be overridden using |
9
|
|
|
* "groupByStrategy" in the query context. If neither the context field nor the property is set, the "v2" strategy |
10
|
|
|
* will be used. |
11
|
|
|
* |
12
|
|
|
* "v2", the default, is designed to offer better performance and memory management. This strategy generates |
13
|
|
|
* per-segment results using a fully off-heap map. Data processes merge the per-segment results using a fully |
14
|
|
|
* off-heap concurrent facts map combined with an on-heap string dictionary. This may optionally involve spilling |
15
|
|
|
* to disk. Data processes return sorted results to the Broker, which merges result streams using an N-way merge. |
16
|
|
|
* The broker materializes the results if necessary (e.g. if the query sorts on columns other than its dimensions). |
17
|
|
|
* Otherwise, it streams results back as they are merged. |
18
|
|
|
* |
19
|
|
|
* "v1", a legacy engine, generates per-segment results on data processes (Historical, realtime, MiddleManager) |
20
|
|
|
* using a map which is partially on-heap (dimension keys and the map itself) and partially off-heap (the |
21
|
|
|
* aggregated values). Data processes then merge the per-segment results using Druid's indexing mechanism. This |
22
|
|
|
* merging is multi-threaded by default, but can optionally be single-threaded. The Broker merges the final result |
23
|
|
|
* set using Druid's indexing mechanism again. The broker merging is always single-threaded. Because the Broker |
24
|
|
|
* merges results using the indexing mechanism, it must materialize the full result set before returning any |
25
|
|
|
* results. On both the data processes and the Broker, the merging index is fully on-heap by default, but it can |
26
|
|
|
* optionally store aggregated values off-heap. |
27
|
|
|
* |
28
|
|
|
* Overrides the value of druid.query.groupBy.defaultStrategy for this query. |
29
|
|
|
*/ |
30
|
|
|
class GroupByV2QueryContext extends QueryContext implements ContextInterface |
31
|
|
|
{ |
32
|
9 |
|
public function __construct(array $properties = []) |
33
|
|
|
{ |
34
|
9 |
|
parent::__construct($properties); |
35
|
|
|
|
36
|
8 |
|
$this->properties['groupByStrategy'] = 'v2'; |
37
|
8 |
|
} |
38
|
|
|
|
39
|
|
|
/** |
40
|
|
|
* Overrides the value of druid.query.groupBy.singleThreaded for this query. |
41
|
|
|
* |
42
|
|
|
* @param bool $groupByIsSingleThreaded |
43
|
|
|
* |
44
|
|
|
* @return $this; |
45
|
|
|
*/ |
46
|
1 |
|
public function setGroupByIsSingleThreaded(bool $groupByIsSingleThreaded) |
47
|
|
|
{ |
48
|
1 |
|
$this->properties['groupByIsSingleThreaded'] = $groupByIsSingleThreaded; |
49
|
|
|
|
50
|
1 |
|
return $this; |
51
|
|
|
} |
52
|
|
|
|
53
|
|
|
/** |
54
|
|
|
* Overrides the value of druid.query.groupBy.bufferGrouperInitialBuckets for this query. |
55
|
|
|
* |
56
|
|
|
* @param int $bufferGrouperInitialBuckets |
57
|
|
|
* |
58
|
|
|
* @return $this; |
59
|
|
|
*/ |
60
|
1 |
|
public function setBufferGrouperInitialBuckets(int $bufferGrouperInitialBuckets) |
61
|
|
|
{ |
62
|
1 |
|
$this->properties['bufferGrouperInitialBuckets'] = $bufferGrouperInitialBuckets; |
63
|
|
|
|
64
|
1 |
|
return $this; |
65
|
|
|
} |
66
|
|
|
|
67
|
|
|
/** |
68
|
|
|
* Overrides the value of druid.query.groupBy.bufferGrouperMaxLoadFactor for this query. |
69
|
|
|
* |
70
|
|
|
* @param int $bufferGrouperMaxLoadFactor |
71
|
|
|
* |
72
|
|
|
* @return $this; |
73
|
|
|
*/ |
74
|
1 |
|
public function setBufferGrouperMaxLoadFactor(int $bufferGrouperMaxLoadFactor) |
75
|
|
|
{ |
76
|
1 |
|
$this->properties['bufferGrouperMaxLoadFactor'] = $bufferGrouperMaxLoadFactor; |
77
|
|
|
|
78
|
1 |
|
return $this; |
79
|
|
|
} |
80
|
|
|
|
81
|
|
|
/** |
82
|
|
|
* Overrides the value of druid.query.groupBy.forceHashAggregation |
83
|
|
|
* |
84
|
|
|
* @param bool $forceHashAggregation |
85
|
|
|
* |
86
|
|
|
* @return $this; |
87
|
|
|
*/ |
88
|
1 |
|
public function setForceHashAggregation(bool $forceHashAggregation) |
89
|
|
|
{ |
90
|
1 |
|
$this->properties['forceHashAggregation'] = $forceHashAggregation; |
91
|
|
|
|
92
|
1 |
|
return $this; |
93
|
|
|
} |
94
|
|
|
|
95
|
|
|
/** |
96
|
|
|
* Overrides the value of druid.query.groupBy.intermediateCombineDegree |
97
|
|
|
* |
98
|
|
|
* @param int $intermediateCombineDegree |
99
|
|
|
* |
100
|
|
|
* @return $this; |
101
|
|
|
*/ |
102
|
1 |
|
public function setIntermediateCombineDegree(int $intermediateCombineDegree) |
103
|
|
|
{ |
104
|
1 |
|
$this->properties['intermediateCombineDegree'] = $intermediateCombineDegree; |
105
|
|
|
|
106
|
1 |
|
return $this; |
107
|
|
|
} |
108
|
|
|
|
109
|
|
|
/** |
110
|
|
|
* Overrides the value of druid.query.groupBy.numParallelCombineThreads |
111
|
|
|
* |
112
|
|
|
* @param int $numParallelCombineThreads |
113
|
|
|
* |
114
|
|
|
* @return $this; |
115
|
|
|
*/ |
116
|
1 |
|
public function setNumParallelCombineThreads(int $numParallelCombineThreads) |
117
|
|
|
{ |
118
|
1 |
|
$this->properties['numParallelCombineThreads'] = $numParallelCombineThreads; |
119
|
|
|
|
120
|
1 |
|
return $this; |
121
|
|
|
} |
122
|
|
|
|
123
|
|
|
/** |
124
|
|
|
* Sort the results first by dimension values and then by timestamp. |
125
|
|
|
* |
126
|
|
|
* @param bool $sortByDimsFirst |
127
|
|
|
* |
128
|
|
|
* @return $this; |
129
|
|
|
*/ |
130
|
1 |
|
public function setSortByDimsFirst(bool $sortByDimsFirst) |
131
|
|
|
{ |
132
|
1 |
|
$this->properties['sortByDimsFirst'] = $sortByDimsFirst; |
133
|
|
|
|
134
|
1 |
|
return $this; |
135
|
|
|
} |
136
|
|
|
|
137
|
|
|
/** |
138
|
|
|
* When all fields in the orderBy are part of the grouping key, the Broker will push limit application down to the |
139
|
|
|
* Historical processes. When the sorting order uses fields that are not in the grouping key, applying this |
140
|
|
|
* optimization can result in approximate results with unknown accuracy, so this optimization is disabled by |
141
|
|
|
* default in that case. Enabling this context flag turns on limit push down for limit/orderBy's that contain |
142
|
|
|
* non-grouping key columns. |
143
|
|
|
* |
144
|
|
|
* @param bool $forceLimitPushDown |
145
|
|
|
* |
146
|
|
|
* @return $this; |
147
|
|
|
*/ |
148
|
1 |
|
public function setForceLimitPushDown(bool $forceLimitPushDown) |
149
|
|
|
{ |
150
|
1 |
|
$this->properties['forceLimitPushDown'] = $forceLimitPushDown; |
151
|
|
|
|
152
|
1 |
|
return $this; |
153
|
|
|
} |
154
|
|
|
|
155
|
|
|
/** |
156
|
|
|
* Can be used to lower the value of druid.query.groupBy.maxMergingDictionarySize for this query. |
157
|
|
|
* |
158
|
|
|
* @param int $maxMergingDictionarySize |
159
|
|
|
* |
160
|
|
|
* @return $this; |
161
|
|
|
*/ |
162
|
1 |
|
public function setMaxMergingDictionarySize(int $maxMergingDictionarySize) |
163
|
|
|
{ |
164
|
1 |
|
$this->properties['maxMergingDictionarySize'] = $maxMergingDictionarySize; |
165
|
|
|
|
166
|
1 |
|
return $this; |
167
|
|
|
} |
168
|
|
|
|
169
|
|
|
/** |
170
|
|
|
* Can be used to lower the value of druid.query.groupBy.maxOnDiskStorage for this query. |
171
|
|
|
* |
172
|
|
|
* @param int $maxOnDiskStorage |
173
|
|
|
* |
174
|
|
|
* @return $this; |
175
|
|
|
*/ |
176
|
3 |
|
public function setMaxOnDiskStorage(int $maxOnDiskStorage) |
177
|
|
|
{ |
178
|
3 |
|
$this->properties['maxOnDiskStorage'] = $maxOnDiskStorage; |
179
|
|
|
|
180
|
3 |
|
return $this; |
181
|
|
|
} |
182
|
|
|
|
183
|
|
|
/** |
184
|
|
|
* If Broker pushes limit down to queryable nodes (historicals, peons) then limit results |
185
|
|
|
* during segment scan. This context value can be used to override |
186
|
|
|
* druid.query.groupBy.applyLimitPushDownToSegment. |
187
|
|
|
* |
188
|
|
|
* @param bool $applyLimitPushDownToSegment |
189
|
|
|
*/ |
190
|
1 |
|
public function setApplyLimitPushDownToSegment(bool $applyLimitPushDownToSegment) |
191
|
|
|
{ |
192
|
1 |
|
$this->properties['applyLimitPushDownToSegment'] = $applyLimitPushDownToSegment; |
193
|
|
|
|
194
|
1 |
|
return $this; |
195
|
|
|
} |
196
|
|
|
} |