Coverage Report

Created: 2026-11-02 14:38

next uncovered line (L), next uncovered region (R), next uncovered branch (B)
be/src/storage/changelog.h
Line
Count
Source
1
// Licensed to the Apache Software Foundation (ASF) under one
2
// or more contributor license agreements.  See the NOTICE file
3
// distributed with this work for additional information
4
// regarding copyright ownership.  The ASF licenses this file
5
// to you under the Apache License, Version 2.0 (the
6
// "License"); you may not use this file except in compliance
7
// with the License.  You may obtain a copy of the License at
8
//
9
//   http://www.apache.org/licenses/LICENSE-2.0
10
//
11
// Unless required by applicable law or agreed to in writing,
12
// software distributed under the License is distributed on an
13
// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
14
// KIND, either express or implied.  See the License for the
15
// specific language governing permissions and limitations
16
// under the License.
17
18
#pragma once
19
20
#include <limits>
21
#include <map>
22
#include <memory>
23
#include <mutex>
24
#include <string>
25
#include <string_view>
26
#include <utility>
27
#include <vector>
28
29
#include "common/logging.h" // DCHECK
30
#include "common/status.h"
31
#include "exec/sink/autoinc_buffer.h"
32
#include "storage/changelog_types.h"
33
#include "storage/olap_common.h"
34
#include "storage/olap_define.h" // DataWriteType
35
#include "storage/rowset/rowset_fwd.h"
36
#include "storage/tablet/tablet_fwd.h"    // BaseTabletSPtr
37
#include "storage/tablet/tablet_schema.h" // TabletSchemaSPtr
38
#include "storage/utils.h"                // CHANGELOG_*_COL
39
40
namespace doris {
41
42
class AutoIncIDBuffer;
43
class PChangelogWriteColumnMappings;
44
class TChangelogWriteColumnMappings;
45
struct PartialUpdateInfo;
46
struct MowContext;
47
48
namespace changelog {
49
50
// A layout belongs to exactly these two pinned schemas. Writers share it for
51
// their operation; no CID is stored in rowset metadata or reused after recovery.
52
struct ResolvedChangelogLayout {
53
    TabletSchemaSPtr source_schema;
54
    TabletSchemaSPtr target_schema;
55
    std::vector<uint32_t> current_source_cids;
56
    std::vector<uint32_t> current_target_cids;
57
    std::vector<bool> current_nullable_widening;
58
    std::vector<uint32_t> before_source_cids;
59
    std::vector<uint32_t> before_target_cids;
60
    uint32_t tso_cid;
61
    uint32_t lsn_cid;
62
    uint32_t op_cid;
63
    bool need_historical_value;
64
};
65
66
Status validate_mapping(const PChangelogWriteColumnMappings& mapping);
67
Status mapping_from_thrift(const TChangelogWriteColumnMappings& input,
68
                           PChangelogWriteColumnMappings* output);
69
Status resolve_layout(const TabletSchemaSPtr& source, const TabletSchemaSPtr& target,
70
                      const PChangelogWriteColumnMappings& mapping,
71
                      std::shared_ptr<const ResolvedChangelogLayout>* result);
72
73
// ATTN(changelog-rename): formerly "__BEFORE__<col>__"
74
293
inline std::string build_before_column_name(std::string_view name) {
75
293
    std::string before_name = "__DORIS_CHANGELOG_BEFORE_";
76
293
    before_name.append(name.data(), name.size());
77
293
    before_name.append("__");
78
293
    return before_name;
79
293
}
80
81
} // namespace changelog
82
83
constexpr int64_t kChangelogLsnAutoIncId = -1;
84
85
// Allocate per-row LSNs.
86
// The caller must provide a valid auto-inc buffer (typically from GlobalAutoIncBuffers).
87
inline Status allocate_lsn(const std::shared_ptr<AutoIncIDBuffer>& lsn_buffer, size_t num_rows,
88
7
                           std::vector<int64_t>& lsn_ids) {
89
7
    if (lsn_buffer == nullptr) {
90
0
        return Status::InternalError("try to get lsn buffer but null");
91
0
    }
92
7
    DCHECK(num_rows > 0);
93
94
7
    std::vector<std::pair<int64_t, size_t>> ranges;
95
7
    RETURN_IF_ERROR(lsn_buffer->sync_request_ids(num_rows, &ranges));
96
97
7
    lsn_ids.clear();
98
7
    lsn_ids.reserve(num_rows);
99
7
    for (const auto& [start, length] : ranges) {
100
18
        for (size_t i = 0; i < length; ++i) {
101
11
            DCHECK_LE(start, std::numeric_limits<int64_t>::max() - static_cast<int64_t>(i));
102
11
            lsn_ids.push_back(start + static_cast<int64_t>(i));
103
11
        }
104
7
    }
105
7
    DCHECK_EQ(lsn_ids.size(), num_rows);
106
7
    return Status::OK();
107
7
}
108
109
constexpr int64_t kTsoLogicalBits = 18;
110
111
29
inline int64_t extract_tso_physical_time(int64_t tso) {
112
29
    return tso <= 0 ? 0 : tso >> kTsoLogicalBits;
113
29
}
114
115
28
inline int64_t changelog_ttl_cutoff_tso(int64_t reference_tso, int64_t ttl_seconds) {
116
28
    DCHECK_GT(reference_tso, 0);
117
28
    DCHECK_GE(ttl_seconds, 0);
118
28
    const int64_t reference_ms = extract_tso_physical_time(reference_tso);
119
28
    if (ttl_seconds > reference_ms / 1000) {
120
2
        return 0;
121
2
    }
122
26
    constexpr int64_t kMaxLogical = (1L << kTsoLogicalBits) - 1;
123
26
    return ((reference_ms - ttl_seconds * 1000) << kTsoLogicalBits) | kMaxLogical;
124
28
}
125
126
namespace segment_v2 {
127
128
class SegmentAllocatedLsnMap {
129
public:
130
68
    void insert_segment_allocated_lsns(int64_t seg_id, ConstAllocatedLsnVectorSharedPtr lsn_ids) {
131
68
        std::lock_guard<std::mutex> l(_mutex);
132
68
        _seg_id_to_lsn_ids.emplace(seg_id, std::move(lsn_ids));
133
68
    }
134
135
54
    void remove_segment(int64_t seg_id) {
136
54
        std::lock_guard<std::mutex> l(_mutex);
137
54
        _seg_id_to_lsn_ids.erase(seg_id);
138
54
    }
139
140
6
    bool contains_segment(int64_t seg_id) const {
141
6
        std::lock_guard<std::mutex> l(_mutex);
142
6
        return _seg_id_to_lsn_ids.count(seg_id) > 0;
143
6
    }
144
145
53
    ConstAllocatedLsnVectorSharedPtr get_segment_allocated_lsns(int64_t seg_id) const {
146
53
        std::lock_guard<std::mutex> l(_mutex);
147
53
        auto it = _seg_id_to_lsn_ids.find(seg_id);
148
53
        DCHECK(it != _seg_id_to_lsn_ids.end())
149
0
                << "SegmentAllocatedLsnMap::get_segment_allocated_lsns missing seg_id=" << seg_id
150
0
                << ", existing_seg_ids=[" << ([&] {
151
0
                       std::string s;
152
0
                       for (const auto& [id, _] : _seg_id_to_lsn_ids) {
153
0
                           if (!s.empty()) {
154
0
                               s.push_back(',');
155
0
                           }
156
0
                           s.append(std::to_string(id));
157
0
                       }
158
0
                       return s;
159
0
                   }())
160
0
                << "]";
161
53
        return it->second;
162
53
    }
163
164
private:
165
    mutable std::mutex _mutex;
166
    std::map<int64_t, ConstAllocatedLsnVectorSharedPtr> _seg_id_to_lsn_ids;
167
};
168
169
struct SegmentWriteChangelogOptions {
170
public:
171
    bool write_before = false;
172
    std::shared_ptr<const changelog::ResolvedChangelogLayout> layout;
173
174
    // source context, used for retrieving historical row and building changelog block
175
    struct SourceWriteDataOptions {
176
        BaseTabletSPtr base_tablet = nullptr;
177
        TabletSchemaSPtr tablet_schema = nullptr;
178
        std::shared_ptr<PartialUpdateInfo> partial_update_info;
179
        std::shared_ptr<MowContext> mow_context;
180
        bool is_transient_rowset_writer = false;
181
        DataWriteType source_write_type = DataWriteType::TYPE_DEFAULT;
182
        // input rowset's changelog rowset, read LSN from it at publish time
183
        RowsetSharedPtr changelog_rowset = nullptr;
184
    } source;
185
};
186
} // namespace segment_v2
187
188
} // namespace doris