Class: Fluent::RedshiftOutput

Inherits:
BufferedOutput
  • Object
show all
Defined in:
lib/fluent/plugin/out_redshift-out.rb

Defined Under Namespace

Classes: MaintenanceError, MaintenanceMonitor, RedshiftConnection, RedshiftError

Constant Summary collapse

NULL_CHAR_FOR_COPY = 
"\\N"
IGNORE_REDSHIFT_ERROR_REGEXP =

ignore load table error. (invalid data format)

/^ERROR:  Load into table '[^']+' failed\./

Instance Method Summary collapse

Constructor Details

#initialize ⇒ RedshiftOutput

Returns a new instance of RedshiftOutput.



12
13
14
15
16
17
18
19
20
21
22
23
 
# File 'lib/fluent/plugin/out_redshift-out.rb', line 12

def initialize
  super
  require 'aws-sdk-v1'
  require 'zlib'
  require 'time'
  require 'tempfile'
  require 'pg'
  require 'csv'
  require 'multi_json'
  require 'yajl'
  ::MultiJson.use(:yajl)
end

Instance Method Details

#configure(conf) ⇒ Object 



53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
 
# File 'lib/fluent/plugin/out_redshift-out.rb', line 53

def configure(conf)
  super
  @path = "#{@path}/" unless @path.end_with?('/') # append last slash
  @path = @path[1..-1] if @path.start_with?('/')  # remove head slash
  @utc = true if conf['utc']
  @db_conf = {
    host:@redshift_host,
    port:@redshift_port,
    dbname:@redshift_dbname,
    user:@redshift_user,
    password:@redshift_password,
    connect_timeout: @redshift_connect_timeout
  }
  @delimiter = determine_delimiter(@file_type) if @delimiter.nil? or @delimiter.empty?
  $log.debug format_log("redshift file_type:#{@file_type} delimiter:'#{@delimiter}'")
  @table_name_with_schema = [@redshift_schemaname, @redshift_tablename].compact.join('.')
  @maintenance_monitor = MaintenanceMonitor.new(@maintenance_file_path)
end

#format(tag, time, record) ⇒ Object 



87
88
89
90
91
92
93
94
95
 
# File 'lib/fluent/plugin/out_redshift-out.rb', line 87

def format(tag, time, record)
  if json?
    record.to_msgpack
  elsif msgpack?
    { @record_log_tag => record }.to_msgpack
  else
    "#{record[@record_log_tag]}\n"
  end
end

#start ⇒ Object 



72
73
74
75
76
77
78
79
80
81
82
83
84
85
 
# File 'lib/fluent/plugin/out_redshift-out.rb', line 72

def start
  super
  # init s3 conf
  options = {}
  if @aws_key_id && @aws_sec_key
    options[:access_key_id] = @aws_key_id
    options[:secret_access_key] = @aws_sec_key
  end
  options[:region] = @s3_region if @s3_region
  @s3 = AWS::S3.new(options)
  @bucket = @s3.buckets[@s3_bucket]
  @redshift_connection = RedshiftConnection.new(@db_conf)
  @redshift_connection.connect_start
end

#write(chunk) ⇒ Object 



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
 
# File 'lib/fluent/plugin/out_redshift-out.rb', line 97

def write(chunk)
  $log.debug format_log("start creating gz.")
  @maintenance_monitor.check_maintenance!

  # create a gz file
  tmp = Tempfile.new("s3-")
  tmp =
    if json? || msgpack?
      create_gz_file_from_structured_data(tmp, chunk, @delimiter)
    else
      create_gz_file_from_flat_data(tmp, chunk)
    end

  # no data -> skip
  unless tmp
    $log.debug format_log("received no valid data. ")
    return false # for debug
  end

  # create a file path with time format
  s3path = create_s3path(@bucket, @path)

  # upload gz to s3
  @bucket.objects[s3path].write(Pathname.new(tmp.path),
                                :acl => :bucket_owner_full_control)

  # close temp file
  tmp.close!

  # copy gz on s3 to redshift
  s3_uri = "s3://#{@s3_bucket}/#{s3path}"
  credentials = @s3.client.credential_provider.credentials
  sql = "copy #{@table_name_with_schema} from '#{s3_uri}'"
  sql += " CREDENTIALS 'aws_access_key_id=#{credentials[:access_key_id]};aws_secret_access_key=#{credentials[:secret_access_key]}"
  sql += ";token=#{credentials[:session_token]}" if credentials[:session_token]
  sql += "' delimiter '#{@delimiter}' GZIP ESCAPE #{@redshift_copy_base_options} #{@redshift_copy_options};"

  $log.debug format_log("start copying. s3_uri=#{s3_uri}")

  begin
    @redshift_connection.exec(sql)
    $log.info format_log("completed copying to redshift. s3_uri=#{s3_uri}")
  rescue RedshiftError => e
    if e.to_s =~ IGNORE_REDSHIFT_ERROR_REGEXP
      $log.error format_log("failed to copy data into redshift due to load error. s3_uri=#{s3_uri}"), :error=>e.to_s
      return false # for debug
    end
    raise e
  end
  true # for debug
end