1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
|
/**
* Copyright (C) 2013 10gen Inc.
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU Affero General Public License, version 3,
* as published by the Free Software Foundation.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Affero General Public License for more details.
*
* You should have received a copy of the GNU Affero General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*
* As a special exception, the copyright holders give permission to link the
* code of portions of this program with the OpenSSL library under certain
* conditions as described in each individual source file and distribute
* linked combinations including the program with the OpenSSL library. You
* must comply with the GNU Affero General Public License in all respects for
* all of the code used other than as permitted herein. If you modify file(s)
* with this exception, you may extend this exception to your version of the
* file(s), but you are not obligated to do so. If you do not wish to do so,
* delete this exception statement from your version. If you delete this
* exception statement from all source files in the program, then also delete
* it in the license file.
*/
#pragma once
#include "mongo/db/diskloc.h"
#include "mongo/db/jsobj.h"
#include "mongo/db/exec/plan_stage.h"
#include "mongo/s/chunk_version.h"
#include "mongo/s/d_state.h"
namespace mongo {
/**
* This stage drops documents that didn't belong to the shard we're executing on at the time of
* construction. This matches the contract for sharded cursorids which guarantees that a
* StaleConfigException will be thrown early or the cursorid for its entire lifetime will return
* documents matching the shard version set on the connection at the time of cursorid creation.
*
* A related system will ensure that the data migrated away from a shard will not be deleted as
* long as there are active queries from before the migration. Currently, "active queries" is
* defined by cursorids so it is important that the metadata used in this stage uses the same
* version as the cursorid. Therefore, you must wrap any Runner using this Stage in a
* ClientCursor during the same lock grab as constructing the Runner.
*
* BEGIN NOTE FROM GREG
*
* There are three sharded query contracts:
*
* 0) Migration commit takes the db lock - i.e. is serialized with writes and reads.
* 1) No data should be returned from a query in ranges of migrations that committed after the
* query started, or from ranges not owned when the query began.
* 2) No migrated data should be removed from a shard while there are queries that were active
* before the migration.
*
* As implementation details, collection metadata is used to determine the ranges of all data
* not actively migrated (or orphaned). CursorIds are currently used to establish "active"
* queries before migration commit.
*
* Combining all this: if a query is started in a db lock and acquires in that (same) lock the
* collection metadata and a cursorId, the query will return results for exactly the ranges in
* the metadata (though of arbitrary staleness). This is the sharded collection query contract.
*
* END NOTE FROM GREG
*
* Preconditions: Child must be fetched. TODO: when covering analysis is in just build doc
* and check that against shard key. See SERVER-5022.
*/
class ShardFilterStage : public PlanStage {
public:
ShardFilterStage(const CollectionMetadataPtr& metadata, WorkingSet* ws, PlanStage* child);
virtual ~ShardFilterStage();
virtual bool isEOF();
virtual StageState work(WorkingSetID* out);
virtual void saveState();
virtual void restoreState(OperationContext* opCtx);
virtual void invalidate(const DiskLoc& dl, InvalidationType type);
virtual std::vector<PlanStage*> getChildren() const;
virtual StageType stageType() const { return STAGE_SHARDING_FILTER; }
virtual PlanStageStats* getStats();
virtual const CommonStats* getCommonStats();
virtual const SpecificStats* getSpecificStats();
static const char* kStageType;
private:
WorkingSet* _ws;
scoped_ptr<PlanStage> _child;
// Stats
CommonStats _commonStats;
ShardingFilterStats _specificStats;
// Note: it is important that this is the metadata from the time this stage is constructed.
// See class comment for details.
const CollectionMetadataPtr _metadata;
};
} // namespace mongo
|
