summaryrefslogtreecommitdiff
path: root/spec/lib/gitlab/database/reindexing_spec.rb
blob: 6575c92e313aeff228a3a27f5fa170bcf2e65ef0 (plain)
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
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
# frozen_string_literal: true

require 'spec_helper'

RSpec.describe Gitlab::Database::Reindexing, feature_category: :database, time_travel_to: '2023-01-07T09:44:07Z' do
  include ExclusiveLeaseHelpers
  include Database::DatabaseHelpers

  describe '.invoke' do
    let(:databases) { Gitlab::Database.database_base_models_with_gitlab_shared }
    let(:databases_count) { databases.count }

    it 'cleans up any leftover indexes' do
      expect(described_class).to receive(:cleanup_leftovers!).exactly(databases_count).times

      described_class.invoke
    end

    context 'when there is an error raised' do
      it 'logs and re-raise' do
        expect(described_class).to receive(:automatic_reindexing).and_raise('Unexpected!')
        expect(Gitlab::AppLogger).to receive(:error)

        expect { described_class.invoke }.to raise_error('Unexpected!')
      end
    end

    context 'when async index creation is enabled' do
      it 'executes async index creation prior to any reindexing actions' do
        stub_feature_flags(database_async_index_creation: true)

        expect(Gitlab::Database::AsyncIndexes).to receive(:create_pending_indexes!).ordered.exactly(databases_count).times
        expect(described_class).to receive(:automatic_reindexing).ordered.exactly(databases_count).times

        described_class.invoke
      end
    end

    context 'when async index creation is disabled' do
      it 'does not execute async index creation' do
        stub_feature_flags(database_async_index_creation: false)

        expect(Gitlab::Database::AsyncIndexes).not_to receive(:create_pending_indexes!)

        described_class.invoke
      end
    end

    it 'executes async index destruction prior to any reindexing actions' do
      expect(Gitlab::Database::AsyncIndexes).to receive(:drop_pending_indexes!).ordered.exactly(databases_count).times
      expect(described_class).to receive(:automatic_reindexing).ordered.exactly(databases_count).times

      described_class.invoke
    end

    context 'calls automatic reindexing' do
      it 'uses all candidate indexes' do
        expect(described_class).to receive(:automatic_reindexing).exactly(databases_count).times

        described_class.invoke
      end

      context 'when explicit database is given' do
        it 'skips other databases' do
          expect(described_class).to receive(:automatic_reindexing).once

          described_class.invoke(Gitlab::Database::PRIMARY_DATABASE_NAME)
        end
      end
    end
  end

  describe '.automatic_reindexing' do
    subject { described_class.automatic_reindexing(maximum_records: limit) }

    let(:limit) { 5 }

    before_all do
      swapout_view_for_table(:postgres_indexes, connection: ApplicationRecord.connection)
    end

    before do
      allow(Gitlab::Database::Reindexing).to receive(:cleanup_leftovers!)
      allow(Gitlab::Database::Reindexing).to receive(:perform_from_queue).and_return(0)
      allow(Gitlab::Database::Reindexing).to receive(:perform_with_heuristic).and_return(0)
    end

    it 'cleans up leftovers, before consuming the queue' do
      expect(Gitlab::Database::Reindexing).to receive(:cleanup_leftovers!).ordered
      expect(Gitlab::Database::Reindexing).to receive(:perform_from_queue).ordered

      subject
    end

    context 'with records in the queue' do
      before do
        create(:reindexing_queued_action)
      end

      context 'with enough records in the queue to reach limit' do
        let(:limit) { 1 }

        it 'does not perform reindexing with heuristic' do
          expect(Gitlab::Database::Reindexing).to receive(:perform_from_queue).and_return(limit)
          expect(Gitlab::Database::Reindexing).not_to receive(:perform_with_heuristic)

          subject
        end
      end

      context 'without enough records in the queue to reach limit' do
        let(:limit) { 2 }

        it 'continues if the queue did not have enough records' do
          expect(Gitlab::Database::Reindexing).to receive(:perform_from_queue).ordered.and_return(1)
          expect(Gitlab::Database::Reindexing).to receive(:perform_with_heuristic).with(maximum_records: 1).ordered

          subject
        end
      end
    end
  end

  describe '.perform_with_heuristic' do
    subject { described_class.perform_with_heuristic(candidate_indexes, maximum_records: limit) }

    let(:limit) { 2 }
    let(:coordinator) { instance_double(Gitlab::Database::Reindexing::Coordinator) }
    let(:index_selection) { instance_double(Gitlab::Database::Reindexing::IndexSelection) }
    let(:candidate_indexes) { double }
    let(:indexes) { [double, double] }

    it 'delegates to Coordinator' do
      expect(Gitlab::Database::Reindexing::IndexSelection).to receive(:new).with(candidate_indexes).and_return(index_selection)
      expect(index_selection).to receive(:take).with(limit).and_return(indexes)

      indexes.each do |index|
        expect(Gitlab::Database::Reindexing::Coordinator).to receive(:new).with(index).and_return(coordinator)
        expect(coordinator).to receive(:perform)
      end

      subject
    end
  end

  describe '.perform_from_queue' do
    subject { described_class.perform_from_queue(maximum_records: limit) }

    before_all do
      swapout_view_for_table(:postgres_indexes, connection: ApplicationRecord.connection)
    end

    let(:limit) { 2 }
    let(:queued_actions) { create_list(:reindexing_queued_action, 3) }
    let(:coordinator) { instance_double(Gitlab::Database::Reindexing::Coordinator) }

    before do
      queued_actions.take(limit).each do |action|
        allow(Gitlab::Database::Reindexing::Coordinator).to receive(:new).with(action.index).and_return(coordinator)
        allow(coordinator).to receive(:perform)
      end
    end

    it 'consumes the queue in order of created_at and applies the limit' do
      queued_actions.take(limit).each do |action|
        expect(Gitlab::Database::Reindexing::Coordinator).to receive(:new).ordered.with(action.index).and_return(coordinator)
        expect(coordinator).to receive(:perform)
      end

      subject
    end

    it 'updates queued action and sets state to done' do
      subject

      queue = queued_actions

      queue.shift(limit).each do |action|
        expect(action.reload.state).to eq('done')
      end

      queue.each do |action|
        expect(action.reload.state).to eq('queued')
      end
    end

    it 'updates queued action upon error and sets state to failed' do
      expect(Gitlab::Database::Reindexing::Coordinator).to receive(:new).ordered.with(queued_actions.first.index).and_return(coordinator)
      expect(coordinator).to receive(:perform).and_raise('something went wrong')

      subject

      states = queued_actions.map(&:reload).map(&:state)

      expect(states).to eq(%w(failed done queued))
    end
  end

  describe '.cleanup_leftovers!' do
    subject(:cleanup_leftovers) { described_class.cleanup_leftovers! }

    let(:model) { Gitlab::Database.database_base_models[Gitlab::Database::PRIMARY_DATABASE_NAME] }
    let(:connection) { model.connection }

    around do |example|
      Gitlab::Database::SharedModel.using_connection(connection) do
        example.run
      end
    end

    before do
      connection.execute(<<~SQL)
        CREATE INDEX foobar_ccnew ON users (id);
        CREATE INDEX foobar_ccnew1 ON users (id);
      SQL
    end

    it 'drops both leftover indexes' do
      expect_query("SET lock_timeout TO '60000ms'")
      expect_query("DROP INDEX CONCURRENTLY IF EXISTS \"public\".\"foobar_ccnew\"")
      expect_query("RESET idle_in_transaction_session_timeout; RESET lock_timeout")
      expect_query("SET lock_timeout TO '60000ms'")
      expect_query("DROP INDEX CONCURRENTLY IF EXISTS \"public\".\"foobar_ccnew1\"")
      expect_query("RESET idle_in_transaction_session_timeout; RESET lock_timeout")

      cleanup_leftovers
    end

    def expect_query(sql)
      expect(connection).to receive(:execute).ordered.with(sql).and_wrap_original do |method, sql|
        method.call(sql.sub(/CONCURRENTLY/, ''))
      end
    end
  end
end