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 <fmt/format.h> |
21 | | |
22 | | #include <limits> |
23 | | #include <map> |
24 | | #include <memory> |
25 | | #include <mutex> |
26 | | #include <string> |
27 | | #include <string_view> |
28 | | #include <utility> |
29 | | #include <vector> |
30 | | |
31 | | #include "common/logging.h" // DCHECK |
32 | | #include "common/status.h" |
33 | | #include "exec/sink/autoinc_buffer.h" |
34 | | #include "storage/olap_common.h" |
35 | | #include "storage/olap_define.h" // DataWriteType |
36 | | #include "storage/rowset/rowset_fwd.h" |
37 | | #include "storage/tablet/tablet_fwd.h" // BaseTabletSPtr |
38 | | #include "storage/tablet/tablet_schema.h" // TabletSchemaSPtr |
39 | | #include "storage/utils.h" // BINLOG_*_COL |
40 | | |
41 | | namespace doris { |
42 | | |
43 | | class AutoIncIDBuffer; |
44 | | struct PartialUpdateInfo; |
45 | | struct MowContext; |
46 | | |
47 | | // Row binlog op type. |
48 | | // NOTE: The value is persisted into row binlog data, so keep it stable. |
49 | | static constexpr int64_t ROW_BINLOG_APPEND = 0; |
50 | | static constexpr int64_t ROW_BINLOG_UPDATE = 1; |
51 | | static constexpr int64_t ROW_BINLOG_DELETE = 2; |
52 | | |
53 | | constexpr std::string_view kBinlogPrefix = "binlog_"; |
54 | | constexpr std::string_view kBinlogMetaPrefix = "binlog_meta_"; |
55 | | constexpr std::string_view kBinlogDataPrefix = "binlog_data_"; |
56 | | |
57 | | namespace binlog { |
58 | | |
59 | 326 | inline std::string build_before_column_name(std::string_view name) { |
60 | 326 | std::string before_name = "__BEFORE__"; |
61 | 326 | before_name.append(name.data(), name.size()); |
62 | 326 | before_name.append("__"); |
63 | 326 | return before_name; |
64 | 326 | } |
65 | | |
66 | | } // namespace binlog |
67 | | |
68 | | constexpr int64_t kBinlogLsnAutoIncId = -1; |
69 | | |
70 | | inline auto make_binlog_meta_key(const std::string_view tablet, int64_t version, |
71 | 2 | const std::string_view rowset) { |
72 | 2 | return fmt::format("{}meta_{}_{:020d}_{}", kBinlogPrefix, tablet, version, rowset); |
73 | 2 | } |
74 | | |
75 | | inline auto make_binlog_meta_key(const std::string_view tablet, const std::string_view version_str, |
76 | 0 | const std::string_view rowset) { |
77 | 0 | // TODO(Drogon): use fmt::format not convert to version_num, only string with length prefix '0' |
78 | 0 | int64_t version = std::atoll(version_str.data()); |
79 | 0 | return make_binlog_meta_key(tablet, version, rowset); |
80 | 0 | } |
81 | | |
82 | | inline auto make_binlog_meta_key(const TabletUid& tablet_uid, int64_t version, |
83 | 2 | const RowsetId& rowset_id) { |
84 | 2 | return make_binlog_meta_key(tablet_uid.to_string(), version, rowset_id.to_string()); |
85 | 2 | } |
86 | | |
87 | 8 | inline auto make_binlog_meta_key_prefix(const TabletUid& tablet_uid) { |
88 | 8 | return fmt::format("{}meta_{}_", kBinlogPrefix, tablet_uid.to_string()); |
89 | 8 | } |
90 | | |
91 | 14 | inline auto make_binlog_meta_key_prefix(const TabletUid& tablet_uid, int64_t version) { |
92 | 14 | return fmt::format("{}meta_{}_{:020d}_", kBinlogPrefix, tablet_uid.to_string(), version); |
93 | 14 | } |
94 | | |
95 | | inline auto make_binlog_data_key(const std::string_view tablet, int64_t version, |
96 | 2 | const std::string_view rowset) { |
97 | 2 | return fmt::format("{}data_{}_{:020d}_{}", kBinlogPrefix, tablet, version, rowset); |
98 | 2 | } |
99 | | |
100 | | inline auto make_binlog_data_key(const std::string_view tablet, const std::string_view version, |
101 | 0 | const std::string_view rowset) { |
102 | 0 | return fmt::format("{}data_{}_{:0>20}_{}", kBinlogPrefix, tablet, version, rowset); |
103 | 0 | } |
104 | | |
105 | | inline auto make_binlog_data_key(const TabletUid& tablet_uid, int64_t version, |
106 | 2 | const RowsetId& rowset_id) { |
107 | 2 | return make_binlog_data_key(tablet_uid.to_string(), version, rowset_id.to_string()); |
108 | 2 | } |
109 | | |
110 | | inline auto make_binlog_data_key(const TabletUid& tablet_uid, int64_t version, |
111 | 0 | const std::string_view rowset_id) { |
112 | 0 | return make_binlog_data_key(tablet_uid.to_string(), version, rowset_id); |
113 | 0 | } |
114 | | |
115 | 0 | inline auto make_binlog_data_key_prefix(const TabletUid& tablet_uid, int64_t version) { |
116 | 0 | return fmt::format("{}data_{}_{:020d}_", kBinlogPrefix, tablet_uid.to_string(), version); |
117 | 0 | } |
118 | | |
119 | 0 | inline auto make_binlog_filename_key(const TabletUid& tablet_uid, const std::string_view version) { |
120 | 0 | return fmt::format("{}meta_{}_{:0>20}_", kBinlogPrefix, tablet_uid.to_string(), version); |
121 | 0 | } |
122 | | |
123 | 0 | inline bool starts_with_binlog_meta(const std::string_view str) { |
124 | 0 | auto prefix = kBinlogMetaPrefix; |
125 | 0 | if (prefix.length() > str.length()) { |
126 | 0 | return false; |
127 | 0 | } |
128 | | |
129 | 0 | return str.compare(0, prefix.length(), prefix) == 0; |
130 | 0 | } |
131 | | |
132 | 1 | inline std::string get_binlog_data_key_from_meta_key(const std::string_view meta_key) { |
133 | | // like "binlog_meta_6943f1585fe834b5-e542c2b83a21d0b7" => "binlog_data-6943f1585fe834b5-e542c2b83a21d0b7" |
134 | 1 | return fmt::format("{}data_{}", kBinlogPrefix, meta_key.substr(kBinlogMetaPrefix.length())); |
135 | 1 | } |
136 | | |
137 | | // Allocate per-row LSNs. |
138 | | // The caller must provide a valid auto-inc buffer (typically from GlobalAutoIncBuffers). |
139 | | inline Status allocate_lsn(const std::shared_ptr<AutoIncIDBuffer>& lsn_buffer, size_t num_rows, |
140 | 7 | std::vector<int64_t>& lsn_ids) { |
141 | 7 | if (lsn_buffer == nullptr) { |
142 | 0 | return Status::InternalError("try to get lsn buffer but null"); |
143 | 0 | } |
144 | 7 | DCHECK(num_rows > 0); |
145 | | |
146 | 7 | std::vector<std::pair<int64_t, size_t>> ranges; |
147 | 7 | RETURN_IF_ERROR(lsn_buffer->sync_request_ids(num_rows, &ranges)); |
148 | | |
149 | 7 | lsn_ids.clear(); |
150 | 7 | lsn_ids.reserve(num_rows); |
151 | 7 | for (const auto& [start, length] : ranges) { |
152 | 18 | for (size_t i = 0; i < length; ++i) { |
153 | 11 | DCHECK_LE(start, std::numeric_limits<int64_t>::max() - static_cast<int64_t>(i)); |
154 | 11 | lsn_ids.push_back(start + static_cast<int64_t>(i)); |
155 | 11 | } |
156 | 7 | } |
157 | 7 | DCHECK_EQ(lsn_ids.size(), num_rows); |
158 | 7 | return Status::OK(); |
159 | 7 | } |
160 | | |
161 | | constexpr int64_t kTsoLogicalBits = 18; |
162 | | |
163 | 29 | inline int64_t extract_tso_physical_time(int64_t tso) { |
164 | 29 | return tso <= 0 ? 0 : tso >> kTsoLogicalBits; |
165 | 29 | } |
166 | | |
167 | 28 | inline int64_t row_binlog_ttl_cutoff_tso(int64_t reference_tso, int64_t ttl_seconds) { |
168 | 28 | DCHECK_GT(reference_tso, 0); |
169 | 28 | DCHECK_GE(ttl_seconds, 0); |
170 | 28 | const int64_t reference_ms = extract_tso_physical_time(reference_tso); |
171 | 28 | if (ttl_seconds > reference_ms / 1000) { |
172 | 2 | return 0; |
173 | 2 | } |
174 | 26 | constexpr int64_t kMaxLogical = (1L << kTsoLogicalBits) - 1; |
175 | 26 | return ((reference_ms - ttl_seconds * 1000) << kTsoLogicalBits) | kMaxLogical; |
176 | 28 | } |
177 | | |
178 | | namespace segment_v2 { |
179 | | |
180 | | class SegmentAllocatedLsnMap { |
181 | | public: |
182 | 66 | void insert_segment_allocated_lsns(int64_t seg_id, ConstAllocatedLsnVectorSharedPtr lsn_ids) { |
183 | 66 | std::lock_guard<std::mutex> l(_mutex); |
184 | 66 | _seg_id_to_lsn_ids.emplace(seg_id, std::move(lsn_ids)); |
185 | 66 | } |
186 | | |
187 | 54 | void remove_segment(int64_t seg_id) { |
188 | 54 | std::lock_guard<std::mutex> l(_mutex); |
189 | 54 | _seg_id_to_lsn_ids.erase(seg_id); |
190 | 54 | } |
191 | | |
192 | 6 | bool contains_segment(int64_t seg_id) const { |
193 | 6 | std::lock_guard<std::mutex> l(_mutex); |
194 | 6 | return _seg_id_to_lsn_ids.count(seg_id) > 0; |
195 | 6 | } |
196 | | |
197 | 53 | ConstAllocatedLsnVectorSharedPtr get_segment_allocated_lsns(int64_t seg_id) const { |
198 | 53 | std::lock_guard<std::mutex> l(_mutex); |
199 | 53 | auto it = _seg_id_to_lsn_ids.find(seg_id); |
200 | 53 | DCHECK(it != _seg_id_to_lsn_ids.end()) |
201 | 0 | << "SegmentAllocatedLsnMap::get_segment_allocated_lsns missing seg_id=" << seg_id |
202 | 0 | << ", existing_seg_ids=[" << ([&] { |
203 | 0 | std::string s; |
204 | 0 | for (const auto& [id, _] : _seg_id_to_lsn_ids) { |
205 | 0 | if (!s.empty()) { |
206 | 0 | s.push_back(','); |
207 | 0 | } |
208 | 0 | s.append(std::to_string(id)); |
209 | 0 | } |
210 | 0 | return s; |
211 | 0 | }()) |
212 | 0 | << "]"; |
213 | 53 | return it->second; |
214 | 53 | } |
215 | | |
216 | | private: |
217 | | mutable std::mutex _mutex; |
218 | | std::map<int64_t, ConstAllocatedLsnVectorSharedPtr> _seg_id_to_lsn_ids; |
219 | | }; |
220 | | |
221 | | struct SegmentWriteBinlogOptions { |
222 | | public: |
223 | | bool write_before = false; |
224 | | |
225 | | // source context, used for retrieving historical row and building binlog<row> block |
226 | | struct SourceWriteDataOptions { |
227 | | BaseTabletSPtr base_tablet = nullptr; |
228 | | TabletSchemaSPtr tablet_schema = nullptr; |
229 | | std::shared_ptr<PartialUpdateInfo> partial_update_info; |
230 | | std::shared_ptr<MowContext> mow_context; |
231 | | bool is_transient_rowset_writer = false; |
232 | | DataWriteType source_write_type = DataWriteType::TYPE_DEFAULT; |
233 | | // input rowset's row-binlog, read LSN from it at publish time |
234 | | RowsetSharedPtr row_binlog_rowset = nullptr; |
235 | | } source; |
236 | | }; |
237 | | } // namespace segment_v2 |
238 | | |
239 | | } // namespace doris |