/root/doris/be/src/olap/full_compaction.cpp
Line | Count | Source (jump to first uncovered line) |
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 | | #include "olap/full_compaction.h" |
19 | | |
20 | | #include <glog/logging.h> |
21 | | #include <time.h> |
22 | | |
23 | | #include <memory> |
24 | | #include <mutex> |
25 | | #include <ostream> |
26 | | #include <shared_mutex> |
27 | | |
28 | | #include "common/config.h" |
29 | | #include "common/status.h" |
30 | | #include "olap/compaction.h" |
31 | | #include "olap/cumulative_compaction_policy.h" |
32 | | #include "olap/olap_common.h" |
33 | | #include "olap/olap_define.h" |
34 | | #include "olap/rowset/beta_rowset.h" |
35 | | #include "olap/rowset/rowset.h" |
36 | | #include "olap/schema_change.h" |
37 | | #include "olap/tablet_meta.h" |
38 | | #include "runtime/thread_context.h" |
39 | | #include "util/debug_points.h" |
40 | | #include "util/thread.h" |
41 | | #include "util/trace.h" |
42 | | |
43 | | namespace doris { |
44 | | using namespace ErrorCode; |
45 | | |
46 | | FullCompaction::FullCompaction(StorageEngine& engine, const TabletSharedPtr& tablet) |
47 | 0 | : CompactionMixin(engine, tablet, "FullCompaction:" + std::to_string(tablet->tablet_id())) { |
48 | 0 | } |
49 | | |
50 | 0 | FullCompaction::~FullCompaction() { |
51 | 0 | tablet()->set_is_full_compaction_running(false); |
52 | 0 | } |
53 | | |
54 | 0 | Status FullCompaction::prepare_compact() { |
55 | 0 | if (!tablet()->init_succeeded()) { |
56 | 0 | return Status::Error<INVALID_ARGUMENT, false>("Full compaction init failed"); |
57 | 0 | } |
58 | | |
59 | 0 | std::unique_lock base_lock(tablet()->get_base_compaction_lock()); |
60 | 0 | std::unique_lock cumu_lock(tablet()->get_cumulative_compaction_lock()); |
61 | 0 | tablet()->set_is_full_compaction_running(true); |
62 | |
|
63 | 0 | DBUG_EXECUTE_IF("FullCompaction.prepare_compact.set_cumu_point", |
64 | 0 | { tablet()->set_cumulative_layer_point(tablet()->max_version_unlocked() + 1); }) |
65 | | |
66 | | // 1. pick rowsets to compact |
67 | 0 | RETURN_IF_ERROR(pick_rowsets_to_compact()); |
68 | | |
69 | 0 | return Status::OK(); |
70 | 0 | } |
71 | | |
72 | 0 | Status FullCompaction::execute_compact() { |
73 | 0 | std::unique_lock base_lock(tablet()->get_base_compaction_lock()); |
74 | 0 | std::unique_lock cumu_lock(tablet()->get_cumulative_compaction_lock()); |
75 | |
|
76 | 0 | SCOPED_ATTACH_TASK(_mem_tracker); |
77 | |
|
78 | 0 | RETURN_IF_ERROR(CompactionMixin::execute_compact()); |
79 | | |
80 | 0 | Version last_version = _input_rowsets.back()->version(); |
81 | 0 | tablet()->cumulative_compaction_policy()->update_cumulative_point(tablet(), _input_rowsets, |
82 | 0 | _output_rowset, last_version); |
83 | 0 | VLOG_CRITICAL << "after cumulative compaction, current cumulative point is " |
84 | 0 | << tablet()->cumulative_layer_point() << ", tablet=" << _tablet->tablet_id(); |
85 | |
|
86 | 0 | tablet()->set_last_full_compaction_success_time(UnixMillis()); |
87 | |
|
88 | 0 | return Status::OK(); |
89 | 0 | } |
90 | | |
91 | 0 | Status FullCompaction::pick_rowsets_to_compact() { |
92 | 0 | _input_rowsets = tablet()->pick_candidate_rowsets_to_full_compaction(); |
93 | 0 | RETURN_IF_ERROR(check_version_continuity(_input_rowsets)); |
94 | 0 | RETURN_IF_ERROR(_check_all_version(_input_rowsets)); |
95 | 0 | if (_input_rowsets.size() <= 1) { |
96 | 0 | return Status::Error<FULL_NO_SUITABLE_VERSION>("There is no suitable version"); |
97 | 0 | } |
98 | | |
99 | 0 | if (_input_rowsets.size() == 2 && _input_rowsets[0]->end_version() == 1) { |
100 | | // the tablet is with rowset: [0-1], [2-y] |
101 | | // and [0-1] has no data. in this situation, no need to do full compaction. |
102 | 0 | return Status::Error<FULL_NO_SUITABLE_VERSION>("There is no suitable version"); |
103 | 0 | } |
104 | | |
105 | 0 | return Status::OK(); |
106 | 0 | } |
107 | | |
108 | 0 | Status FullCompaction::modify_rowsets() { |
109 | 0 | std::vector<RowsetSharedPtr> output_rowsets {_output_rowset}; |
110 | 0 | if (_tablet->keys_type() == KeysType::UNIQUE_KEYS && |
111 | 0 | _tablet->enable_unique_key_merge_on_write()) { |
112 | 0 | std::vector<RowsetSharedPtr> tmp_rowsets {}; |
113 | | |
114 | | // tablet is under alter process. The delete bitmap will be calculated after conversion. |
115 | 0 | if (_tablet->tablet_state() == TABLET_NOTREADY) { |
116 | 0 | LOG(INFO) << "tablet is under alter process, update delete bitmap later, tablet_id=" |
117 | 0 | << _tablet->tablet_id(); |
118 | 0 | return Status::OK(); |
119 | 0 | } |
120 | | |
121 | 0 | int64_t max_version = tablet()->max_version().second; |
122 | 0 | DCHECK(max_version >= _output_rowset->version().second); |
123 | 0 | if (max_version > _output_rowset->version().second) { |
124 | 0 | RETURN_IF_ERROR(_tablet->capture_consistent_rowsets_unlocked( |
125 | 0 | {_output_rowset->version().second + 1, max_version}, &tmp_rowsets)); |
126 | 0 | } |
127 | | |
128 | 0 | for (const auto& it : tmp_rowsets) { |
129 | 0 | const int64_t& cur_version = it->rowset_meta()->start_version(); |
130 | 0 | RETURN_IF_ERROR(_full_compaction_calc_delete_bitmap(it, _output_rowset, cur_version, |
131 | 0 | _output_rs_writer.get())); |
132 | 0 | } |
133 | | |
134 | 0 | DBUG_EXECUTE_IF("FullCompaction.modify_rowsets.before.block", DBUG_BLOCK); |
135 | 0 | std::lock_guard rowset_update_lock(tablet()->get_rowset_update_lock()); |
136 | 0 | std::lock_guard header_lock(_tablet->get_header_lock()); |
137 | 0 | SCOPED_SIMPLE_TRACE_IF_TIMEOUT(TRACE_TABLET_LOCK_THRESHOLD); |
138 | 0 | for (const auto& it : tablet()->rowset_map()) { |
139 | 0 | const int64_t& cur_version = it.first.first; |
140 | 0 | const RowsetSharedPtr& published_rowset = it.second; |
141 | 0 | if (cur_version > max_version) { |
142 | 0 | RETURN_IF_ERROR(_full_compaction_calc_delete_bitmap( |
143 | 0 | published_rowset, _output_rowset, cur_version, _output_rs_writer.get())); |
144 | 0 | } |
145 | 0 | } |
146 | 0 | RETURN_IF_ERROR(tablet()->modify_rowsets(output_rowsets, _input_rowsets, true)); |
147 | 0 | DBUG_EXECUTE_IF("FullCompaction.modify_rowsets.sleep", { sleep(5); }) |
148 | 0 | tablet()->save_meta(); |
149 | 0 | } else { |
150 | 0 | DBUG_EXECUTE_IF("FullCompaction.modify_rowsets.before.block", DBUG_BLOCK); |
151 | 0 | std::lock_guard<std::mutex> rowset_update_wlock(tablet()->get_rowset_update_lock()); |
152 | 0 | std::lock_guard<std::shared_mutex> meta_wlock(_tablet->get_header_lock()); |
153 | 0 | RETURN_IF_ERROR(tablet()->modify_rowsets(output_rowsets, _input_rowsets, true)); |
154 | 0 | DBUG_EXECUTE_IF("FullCompaction.modify_rowsets.sleep", { sleep(5); }) |
155 | 0 | tablet()->save_meta(); |
156 | 0 | } |
157 | 0 | return Status::OK(); |
158 | 0 | } |
159 | | |
160 | 0 | Status FullCompaction::_check_all_version(const std::vector<RowsetSharedPtr>& rowsets) { |
161 | 0 | if (rowsets.empty()) { |
162 | 0 | return Status::Error<FULL_MISS_VERSION, false>( |
163 | 0 | "There is no input rowset when do full compaction"); |
164 | 0 | } |
165 | 0 | const RowsetSharedPtr& last_rowset = rowsets.back(); |
166 | 0 | const RowsetSharedPtr& first_rowset = rowsets.front(); |
167 | 0 | auto max_version = tablet()->max_version(); |
168 | 0 | if (last_rowset->version() != max_version || first_rowset->version().first != 0) { |
169 | 0 | return Status::Error<FULL_MISS_VERSION, false>( |
170 | 0 | "Full compaction rowsets' versions not equal to all exist rowsets' versions. " |
171 | 0 | "full compaction rowsets max version={}-{}" |
172 | 0 | ", current rowsets max version={}-{}" |
173 | 0 | ", full compaction rowsets min version={}-{}, current rowsets min version=0-1", |
174 | 0 | last_rowset->start_version(), last_rowset->end_version(), max_version.first, |
175 | 0 | max_version.second, first_rowset->start_version(), first_rowset->end_version()); |
176 | 0 | } |
177 | 0 | return Status::OK(); |
178 | 0 | } |
179 | | |
180 | | Status FullCompaction::_full_compaction_calc_delete_bitmap(const RowsetSharedPtr& published_rowset, |
181 | | const RowsetSharedPtr& rowset, |
182 | | int64_t cur_version, |
183 | 0 | RowsetWriter* rowset_writer) { |
184 | 0 | std::vector<segment_v2::SegmentSharedPtr> segments; |
185 | 0 | RETURN_IF_ERROR( |
186 | 0 | std::static_pointer_cast<BetaRowset>(published_rowset)->load_segments(&segments)); |
187 | 0 | DeleteBitmapPtr delete_bitmap = |
188 | 0 | std::make_shared<DeleteBitmap>(_tablet->tablet_meta()->tablet_id()); |
189 | 0 | std::vector<RowsetSharedPtr> specified_rowsets {rowset}; |
190 | |
|
191 | 0 | OlapStopWatch watch; |
192 | 0 | RETURN_IF_ERROR(BaseTablet::calc_delete_bitmap(_tablet, published_rowset, segments, |
193 | 0 | specified_rowsets, delete_bitmap, cur_version, |
194 | 0 | nullptr, rowset_writer)); |
195 | 0 | size_t total_rows = std::accumulate( |
196 | 0 | segments.begin(), segments.end(), 0, |
197 | 0 | [](size_t sum, const segment_v2::SegmentSharedPtr& s) { return sum += s->num_rows(); }); |
198 | 0 | for (const auto& [k, v] : delete_bitmap->delete_bitmap) { |
199 | 0 | if (std::get<1>(k) != DeleteBitmap::INVALID_SEGMENT_ID) { |
200 | 0 | _tablet->tablet_meta()->delete_bitmap().merge( |
201 | 0 | {std::get<0>(k), std::get<1>(k), cur_version}, v); |
202 | 0 | } |
203 | 0 | } |
204 | 0 | VLOG_DEBUG << "[Full compaction] construct delete bitmap tablet: " << _tablet->tablet_id() |
205 | 0 | << ", published rowset version: [" << published_rowset->version().first << "-" |
206 | 0 | << published_rowset->version().second << "]" |
207 | 0 | << ", full compaction rowset version: [" << rowset->version().first << "-" |
208 | 0 | << rowset->version().second << "]" |
209 | 0 | << ", cost: " << watch.get_elapse_time_us() << "(us), total rows: " << total_rows; |
210 | 0 | return Status::OK(); |
211 | 0 | } |
212 | | |
213 | | } // namespace doris |