Class: Fluent::Plugin::MysqlReplicatorMultiInput

Inherits:
Input
  • Object
show all
Defined in:
lib/fluent/plugin/in_mysql_replicator_multi.rb

Instance Method Summary collapse

Constructor Details

#initialize ⇒ MysqlReplicatorMultiInput

Returns a new instance of MysqlReplicatorMultiInput.



7
8
9
10
11
 
# File 'lib/fluent/plugin/in_mysql_replicator_multi.rb', line 7

def initialize
  require 'mysql2'
  require 'digest/sha1'
  super
end

Instance Method Details

#add_hash_table_buffer(setting_name, id, hash) ⇒ Object 



210
211
212
213
 
# File 'lib/fluent/plugin/in_mysql_replicator_multi.rb', line 210

def add_hash_table_buffer(setting_name, id, hash)
  @hash_table_bulk_insert << "('#{setting_name}','#{id}','#{hash}')"
  flush_hash_table if @hash_table_bulk_insert.size >= @bulk_insert_count
end

#collect_gap_ids(config, current_id, previous_id) ⇒ Object 



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
 
# File 'lib/fluent/plugin/in_mysql_replicator_multi.rb', line 164

def collect_gap_ids(config, current_id, previous_id)
  setting_name = config['name']
  if (current_id - previous_id) > 1 && config['enable_loose_delete'] == 0
    query = "SELECT SQL_NO_CACHE setting_query_pk FROM hash_tables
      WHERE setting_name = '#{setting_name}'
      AND setting_query_pk > #{previous_id.to_i} AND setting_query_pk < #{current_id.to_i}"
  elsif (current_id - previous_id) > 1 && config['enable_loose_delete'] == 1
    return [*previous_id...current_id] - [current_id,previous_id]
  elsif previous_id > current_id
    query = "SELECT SQL_NO_CACHE setting_query_pk FROM hash_tables
      WHERE setting_name = '#{setting_name}'
      AND setting_query_pk > #{previous_id.to_i}"
  elsif previous_id == current_id
    query = "SELECT SQL_NO_CACHE setting_query_pk FROM hash_tables
      WHERE setting_name = '#{setting_name}'
      AND (setting_query_pk > #{current_id.to_i} OR setting_query_pk < #{current_id.to_i})"
  end
  ids = Array.new
  unless query.nil?
    @manager_db.query(query).each do |row|
      ids << row['setting_query_pk']
    end
  end
  return ids
end

#configure(conf) ⇒ Object 



23
24
25
26
27
28
 
# File 'lib/fluent/plugin/in_mysql_replicator_multi.rb', line 23

def configure(conf)
  super
  if @tag.nil?
    raise Fluent::ConfigError, "mysql_replicator_multi: missing 'tag' parameter. Please add following line into config like 'tag replicator.${name}.${event}.${primary_key}'"
  end
end

#detect_delete(config, current_id, previous_id) ⇒ Object 



151
152
153
154
155
156
157
158
159
160
161
162
 
# File 'lib/fluent/plugin/in_mysql_replicator_multi.rb', line 151

def detect_delete(config, current_id, previous_id)
  return if config['enable_delete'] != 1 || previous_id.nil?
  deleted_ids = collect_gap_ids(config, current_id, previous_id)
  unless deleted_ids.empty?
    event = :delete
    deleted_ids.each do |id|
      tag = format_tag(@tag, {:name => config['name'], :event => event, :primary_key => config['primary_key']})
      emit_record(tag, {config['primary_key'] => id})
    end
    update_hashtable({:event =>  event, :ids => deleted_ids, :setting_name => config['name']})
  end
end

#detect_insert_update(config, row) ⇒ Object 



126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
 
# File 'lib/fluent/plugin/in_mysql_replicator_multi.rb', line 126

def detect_insert_update(config, row)
  primary_key = config['primary_key']
  current_id = row[primary_key]
  stored_hash = config['enable_loose_insert'] == 1 ? "" : get_stored_hash(config['name'], current_id)
  current_hash = Digest::SHA1.hexdigest(row.flatten.join)

  event = nil
  if stored_hash.empty?
    event = :insert
  elsif stored_hash != current_hash
    event = :update
  end
  unless event.nil?
    tag = format_tag(@tag, {:name => config['name'], :event => event, :primary_key => config['primary_key']})
    emit_record(tag, row)
    update_hashtable({:event => event, :id => current_id, :setting_name => config['name'], :hash => current_hash})
  end
end

#emit_record(tag, record) ⇒ Object 



245
246
247
 
# File 'lib/fluent/plugin/in_mysql_replicator_multi.rb', line 245

def emit_record(tag, record)
  router.emit(tag, Fluent::Engine.now, record)
end

#flush_hash_table ⇒ Object 



235
236
237
238
239
240
241
242
243
 
# File 'lib/fluent/plugin/in_mysql_replicator_multi.rb', line 235

def flush_hash_table
  return if @hash_table_bulk_insert.empty?
  query = "INSERT INTO hash_tables (setting_name,setting_query_pk,setting_query_hash)
    VALUES #{@hash_table_bulk_insert.join(',')}
    ON DUPLICATE KEY UPDATE setting_query_hash = VALUES(setting_query_hash)"
  @manager_db.query(query)
  @hash_table_bulk_insert.clear
  @hash_table_bulk_insert_last_time = Time.now
end

#format_tag(tag, param) ⇒ Object 



202
203
204
205
206
207
208
 
# File 'lib/fluent/plugin/in_mysql_replicator_multi.rb', line 202

def format_tag(tag, param)
  pattern = {'${name}' => param[:name], '${event}' => param[:event].to_s, '${primary_key}' => param[:primary_key]}
  tag.gsub(/(\${[a-z_]+})/) do
    log.warn "mysql_replicator_multi: unknown placeholder found. :tag=>#{tag} :placeholder=>#{$1}" unless pattern.include?($1)
    pattern[$1]
  end
end

#get_manager_connection ⇒ Object 



249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
 
# File 'lib/fluent/plugin/in_mysql_replicator_multi.rb', line 249

def get_manager_connection
  begin
    return Mysql2::Client.new(
      :host => @manager_host,
      :port => @manager_port,
      :username => @manager_username,
      :password => @manager_password,
      :database => @manager_database,
      :encoding => 'utf8',
      :reconnect => true,
      :stream => false,
      :cache_rows => false
    )
  rescue Mysql2::Error => e
    raise "mysql_replicator_multi: #{e}"
  end
end

#get_origin_connection(config) ⇒ Object 



267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
 
# File 'lib/fluent/plugin/in_mysql_replicator_multi.rb', line 267

def get_origin_connection(config)
  begin
    return Mysql2::Client.new(
      :host => config['host'],
      :port => config['port'],
      :username => config['username'],
      :password => config['password'],
      :database => config['database'],
      :encoding => 'utf8',
      :reconnect => true,
      :stream => true,
      :cache_rows => false
    )
  rescue Mysql2::Error => e
    raise "mysql_replicator_multi: #{e}"
  end
end

#get_settings ⇒ Object 



58
59
60
61
62
63
64
65
66
 
# File 'lib/fluent/plugin/in_mysql_replicator_multi.rb', line 58

def get_settings
  manager_db = get_manager_connection
  settings = []
  query = "SELECT * FROM settings WHERE is_active = 1;"
  manager_db.query(query).each do |row|
    settings << row
  end
  return settings
end

#get_stored_hash(setting_name, id) ⇒ Object 



145
146
147
148
149
 
# File 'lib/fluent/plugin/in_mysql_replicator_multi.rb', line 145

def get_stored_hash(setting_name, id)
  query = "SELECT SQL_NO_CACHE setting_query_hash FROM hash_tables WHERE setting_query_pk = #{id.to_i} AND setting_name = '#{setting_name}'"
  row = @manager_db.query(query).first
  return row.nil? ? "" : row['setting_query_hash']
end

#hash_table_flusher ⇒ Object 



215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
 
# File 'lib/fluent/plugin/in_mysql_replicator_multi.rb', line 215

def hash_table_flusher
  begin
    loop do
      if @hash_table_bulk_insert.empty? || @bulk_insert_timeout > (Time.now - @hash_table_bulk_insert_last_time)
        sleep @bulk_insert_timeout
        next
      end
      @mutex.synchronize {
        flush_hash_table
      }
    end
  rescue StandardError => e
    @mutex.synchronize {
      log.error "mysql_replicator_multi: failed to flush buffered query. :config=>#{masked_config}"
      log.error "error: #{e.message}"
      log.error e.backtrace.join("\n")
    }
  end
end

#poll(config) ⇒ Object 



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
 
# File 'lib/fluent/plugin/in_mysql_replicator_multi.rb', line 68

def poll(config)
  begin
    masked_config = Hash[config.map {|k,v| (k == 'password') ? [k, v.to_s.gsub(/./, '*')] : [k,v]}]
    @mutex.synchronize {
      log.info "mysql_replicator_multi: polling start. :config=>#{masked_config}"
    }
    primary_key = config['primary_key']
    previous_id = current_id = nil
    loop do
      rows_count = 0
      start_time = Time.now
      unless config['prepared_query'].nil?
        nest_db = get_origin_connection(config)
        config['prepared_query'].strip.split(/;/).each do |query|
          nest_db.query(query)
        end
      end
      db = get_origin_connection(config)
      db.query(config['query']).each do |row|
        row.each {|k, v| row[k] = v.to_s if v.is_a?(Time) || v.is_a?(Date) || v.is_a?(BigDecimal)}
        row.select {|k, v| v.to_s.strip.match(/^SELECT[^\$]+\$\{[^\}]+\}/i) }.each do |k, v|
          row[k] = [] unless row[k].is_a?(Array)
          nest_db.query(v.gsub(/\$\{([^\}]+)\}/) {|matched| row[$1].to_s}).each do |nest_row|
            nest_row.each {|k, v| nest_row[k] = v.to_s if v.is_a?(Time) || v.is_a?(Date) || v.is_a?(BigDecimal)}
            row[k] << nest_row
          end
        end
        current_id = row[primary_key]
        @mutex.synchronize {
          if row[primary_key].nil?
            log.error "mysql_replicator_multi: missing primary_key. :setting_name=>#{config['name']} :primary_key=>#{primary_key}"
            break
          end
          detect_insert_update(config, row)
          detect_delete(config, current_id, previous_id)
        }
        previous_id = current_id
        rows_count += 1
      end
      db.close
      unless config['prepared_query'].nil?
        nest_db.close
      end
      elapsed_time = sprintf("%0.02f", Time.now - start_time)
      @mutex.synchronize {
        log.info "mysql_replicator_multi: execution finished. :setting_name=>#{config['name']} :rows_count=>#{rows_count} :elapsed_time=>#{elapsed_time} sec"
      }
      sleep config['interval']
    end
  rescue StandardError => e
    @mutex.synchronize {
      log.error "mysql_replicator_multi: failed to execute query. :config=>#{masked_config}"
      log.error "error: #{e.message}"
      log.error e.backtrace.join("\n")
    }
  end
end

#shutdown ⇒ Object 



54
55
56
 
# File 'lib/fluent/plugin/in_mysql_replicator_multi.rb', line 54

def shutdown
  super
end

#start ⇒ Object 



30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
 
# File 'lib/fluent/plugin/in_mysql_replicator_multi.rb', line 30

def start
  super
  begin
    @threads = []
    @mutex = Mutex.new
    @manager_db = get_manager_connection
    @manager_db.query("SET SESSION wait_timeout=1800;")
    @threads << thread_create(:in_mysql_replicator_flusher) {
      @hash_table_bulk_insert = []
      @hash_table_bulk_insert_last_time = Time.now
      hash_table_flusher
    }
    get_settings.each_with_index do |config, idx|
      @threads << thread_create(:"in_mysql_replicator_pollers_#{idx}") {
        poll(config)
      }
    end
    log.error "mysql_replicator_multi: stop working due to empty configuration" if @threads.empty?
  rescue StandardError => e
    log.error "error: #{e.message}"
    log.error e.backtrace.join("\n")
  end
end

#update_hashtable(opts) ⇒ Object 



190
191
192
193
194
195
196
197
198
199
200
 
# File 'lib/fluent/plugin/in_mysql_replicator_multi.rb', line 190

def update_hashtable(opts)
  case opts[:event]
  when :insert
    add_hash_table_buffer(opts[:setting_name], opts[:id], opts[:hash])
  when :update
    query = "UPDATE hash_tables SET setting_query_hash = '#{opts[:hash]}' WHERE setting_name = '#{opts[:setting_name]}' AND setting_query_pk = '#{opts[:id]}'"
  when :delete
    query = "DELETE FROM hash_tables WHERE setting_name = '#{opts[:setting_name]}' AND setting_query_pk IN(#{opts[:ids].join(',')})"
  end
  @manager_db.query(query) unless query.nil?
end