fluent-plugin-redshift-out2 0.4.1
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/.gitignore +19 -0
- data/Gemfile +3 -0
- data/README.md +156 -0
- data/Rakefile +16 -0
- data/VERSION +1 -0
- data/fluent-plugin-redshift-out.gemspec +27 -0
- data/lib/fluent/plugin/out_redshift-out.rb +397 -0
- data/test/plugin/test_out_redshift.rb +503 -0
- data/test/test_helper.rb +8 -0
- metadata +171 -0
checksums.yaml
ADDED
@@ -0,0 +1,7 @@
|
|
1
|
+
---
|
2
|
+
SHA256:
|
3
|
+
metadata.gz: 1041bb98af3d4cd15dbe1d4886a31b68bc2c1691771247b589373792429608a7
|
4
|
+
data.tar.gz: d8162bac6c674de9760c16c4f2f6ce381c13d6d3a9b52ea7fa170fba11d21cb0
|
5
|
+
SHA512:
|
6
|
+
metadata.gz: a28646dd88bc2b290bc38ae8860e49f7f841c0a4c24a78f0df98b5c41d2f82919ba5dafe25b5b6d3d16ce66b4d7515792fe506406940f026afe209027e4c94ef
|
7
|
+
data.tar.gz: 3a421e6727ee9f020f8a6a257ed3d966c611d003daa17ed88d288c210a497d3e9317f5fa49dd80b2339577e0b089280e0a2b578404c3e8007a98a503136bb611
|
data/.gitignore
ADDED
data/Gemfile
ADDED
data/README.md
ADDED
@@ -0,0 +1,156 @@
|
|
1
|
+
Amazon Redshift output plugin for Fluentd
|
2
|
+
========
|
3
|
+
|
4
|
+
## Overview
|
5
|
+
|
6
|
+
Amazon Redshift output plugin uploads event logs to an Amazon Redshift Cluster. Supportted data formats are csv, tsv and json. An S3 bucket and a Redshift Cluster are required to use this plugin.
|
7
|
+
|
8
|
+
## Installation
|
9
|
+
|
10
|
+
gem install fluent-plugin-redshift-out
|
11
|
+
|
12
|
+
## Configuration
|
13
|
+
|
14
|
+
Format:
|
15
|
+
|
16
|
+
<match my.tag>
|
17
|
+
type redshift-out
|
18
|
+
|
19
|
+
# s3 (for copying data to redshift)
|
20
|
+
aws_key_id YOUR_AWS_KEY_ID
|
21
|
+
aws_sec_key YOUR_AWS_SECRET_KEY
|
22
|
+
s3_bucket YOUR_S3_BUCKET
|
23
|
+
s3_endpoint YOUR_S3_BUCKET_END_POINT
|
24
|
+
path YOUR_S3_PATH
|
25
|
+
timestamp_key_format year=%Y/month=%m/day=%d/hour=%H/%Y%m%d-%H%M
|
26
|
+
|
27
|
+
# redshift
|
28
|
+
redshift_host YOUR_AMAZON_REDSHIFT_CLUSTER_END_POINT
|
29
|
+
redshift_port YOUR_AMAZON_REDSHIFT_CLUSTER_PORT
|
30
|
+
redshift_dbname YOUR_AMAZON_REDSHIFT_CLUSTER_DATABASE_NAME
|
31
|
+
redshift_user YOUR_AMAZON_REDSHIFT_CLUSTER_USER_NAME
|
32
|
+
redshift_password YOUR_AMAZON_REDSHIFT_CLUSTER_PASSWORD
|
33
|
+
redshift_schemaname YOUR_AMAZON_REDSHIFT_CLUSTER_TARGET_SCHEMA_NAME
|
34
|
+
redshift_tablename YOUR_AMAZON_REDSHIFT_CLUSTER_TARGET_TABLE_NAME
|
35
|
+
file_type [tsv|csv|json|msgpack]
|
36
|
+
|
37
|
+
# buffer
|
38
|
+
buffer_type file
|
39
|
+
buffer_path /var/log/fluent/redshift
|
40
|
+
flush_interval 15m
|
41
|
+
buffer_chunk_limit 1g
|
42
|
+
</match>
|
43
|
+
|
44
|
+
Example (watch and upload json formatted apache log):
|
45
|
+
|
46
|
+
<source>
|
47
|
+
type tail
|
48
|
+
path redshift_test.json
|
49
|
+
pos_file redshift_test_json.pos
|
50
|
+
tag redshift.json
|
51
|
+
format /^(?<log>.*)$/
|
52
|
+
</source>
|
53
|
+
|
54
|
+
<match redshift.json>
|
55
|
+
type redshift-out
|
56
|
+
|
57
|
+
# s3 (for copying data to redshift)
|
58
|
+
aws_key_id YOUR_AWS_KEY_ID
|
59
|
+
aws_sec_key YOUR_AWS_SECRET_KEY
|
60
|
+
s3_bucket hapyrus-example
|
61
|
+
s3_endpoint s3.amazonaws.com
|
62
|
+
path path/on/s3/apache_json_log/
|
63
|
+
timestamp_key_format year=%Y/month=%m/day=%d/hour=%H/%Y%m%d-%H%M
|
64
|
+
|
65
|
+
# redshift
|
66
|
+
redshift_host xxx-yyy-zzz.xxxxxxxxxx.us-east-1.redshift.amazonaws.com
|
67
|
+
redshift_port 5439
|
68
|
+
redshift_dbname fluent-redshift-test
|
69
|
+
redshift_user fluent
|
70
|
+
redshift_password fluent-password
|
71
|
+
redshift_tablename apache_log
|
72
|
+
file_type json
|
73
|
+
|
74
|
+
# buffer
|
75
|
+
buffer_type file
|
76
|
+
buffer_path /var/log/fluent/redshift
|
77
|
+
flush_interval 15m
|
78
|
+
buffer_chunk_limit 1g
|
79
|
+
<match>
|
80
|
+
|
81
|
+
+ `type` (required) : The value must be `redshift-out`.
|
82
|
+
|
83
|
+
+ `aws_key_id` : AWS access key id to access s3 bucket.
|
84
|
+
|
85
|
+
+ `aws_sec_key` : AWS securet key id to access s3 bucket.
|
86
|
+
|
87
|
+
+ `s3_bucket` (required) : s3 bucket name. S3 bucket must be same as the region of your Redshift cluster.
|
88
|
+
|
89
|
+
+ `s3_endpoint` : s3 endpoint.
|
90
|
+
|
91
|
+
+ `path` (required) : s3 path to input.
|
92
|
+
|
93
|
+
+ `timestamp_key_format` : The format of the object keys. It can include date-format directives.
|
94
|
+
|
95
|
+
- Default parameter is "year=%Y/month=%m/day=%d/hour=%H/%Y%m%d-%H%M"
|
96
|
+
- For example, the s3 path is as following with the above example configration.
|
97
|
+
<pre>
|
98
|
+
hapyrus-example/apache_json_log/year=2013/month=03/day=05/hour=12/20130305_1215_00.gz
|
99
|
+
hapyrus-example/apache_json_log/year=2013/month=03/day=05/hour=12/20130305_1230_00.gz
|
100
|
+
</pre>
|
101
|
+
|
102
|
+
+ `redshift_host` (required) : the end point(or hostname) of your Amazon Redshift cluster.
|
103
|
+
|
104
|
+
+ `redshift_port` (required) : port number.
|
105
|
+
|
106
|
+
+ `redshift_dbname` (required) : database name.
|
107
|
+
|
108
|
+
+ `redshift_user` (required) : user name.
|
109
|
+
|
110
|
+
+ `redshift_password` (required) : password for the user name.
|
111
|
+
|
112
|
+
+ `redshift_tablename` (required) : table name to store data.
|
113
|
+
|
114
|
+
+ `redshift_schemaname` : schema name to store data. By default, this option is not set and find table without schema as your own search_path.
|
115
|
+
|
116
|
+
+ `redshift_connect_timeout` : maximum time to wait for connection to succeed.
|
117
|
+
|
118
|
+
+ `file_type` : file format of the source data. `csv`, `tsv`, `msgpack` or `json` are available.
|
119
|
+
|
120
|
+
+ `delimiter` : delimiter of the source data. This option will be ignored if `file_type` is specified.
|
121
|
+
|
122
|
+
+ `buffer_type` : buffer type.
|
123
|
+
|
124
|
+
+ `buffer_path` : path prefix of the files to buffer logs.
|
125
|
+
|
126
|
+
+ `flush_interval` : flush interval.
|
127
|
+
|
128
|
+
+ `buffer_chunk_limit` : limit buffer size to chunk.
|
129
|
+
|
130
|
+
+ `utc` : utc time zone. This parameter affects `timestamp_key_format`.
|
131
|
+
|
132
|
+
## Logging examples
|
133
|
+
```ruby
|
134
|
+
# examples by fluent-logger
|
135
|
+
require 'fluent-logger'
|
136
|
+
log = Fluent::Logger::FluentLogger.new(nil, :host => 'localhost', :port => 24224)
|
137
|
+
|
138
|
+
# file_type: csv
|
139
|
+
log.post('your.tag', :log => "12345,12345")
|
140
|
+
|
141
|
+
# file_type: tsv
|
142
|
+
log.post('your.tag', :log => "12345\t12345")
|
143
|
+
|
144
|
+
# file_type: json
|
145
|
+
require 'json'
|
146
|
+
log.post('your.tag', :log => { :user_id => 12345, :data_id => 12345 }.to_json)
|
147
|
+
|
148
|
+
# file_type: msgpack
|
149
|
+
log.post('your.tag', :user_id => 12345, :data_id => 12345)
|
150
|
+
```
|
151
|
+
|
152
|
+
## License
|
153
|
+
|
154
|
+
Copyright (c) 2013 [Hapyrus Inc](http://hapyrus.com)
|
155
|
+
|
156
|
+
[Apache License, Version 2.0](http://www.apache.org/licenses/LICENSE-2.0)
|
data/Rakefile
ADDED
@@ -0,0 +1,16 @@
|
|
1
|
+
require "bundler"
|
2
|
+
Bundler::GemHelper.install_tasks
|
3
|
+
require 'rake/testtask'
|
4
|
+
|
5
|
+
Rake::TestTask.new(:test) do |test|
|
6
|
+
test.libs << 'lib' << 'test'
|
7
|
+
test.test_files = FileList['test/plugin/*.rb']
|
8
|
+
test.verbose = true
|
9
|
+
end
|
10
|
+
|
11
|
+
task :coverage do |t|
|
12
|
+
ENV['COVERAGE'] = '1'
|
13
|
+
Rake::Task["test"].invoke
|
14
|
+
end
|
15
|
+
|
16
|
+
task :default => [:build]
|
data/VERSION
ADDED
@@ -0,0 +1 @@
|
|
1
|
+
0.4.1
|
@@ -0,0 +1,27 @@
|
|
1
|
+
# -*- encoding: utf-8 -*-
|
2
|
+
$:.push File.expand_path('../lib', __FILE__)
|
3
|
+
|
4
|
+
Gem::Specification.new do |gem|
|
5
|
+
gem.name = "fluent-plugin-redshift-out2"
|
6
|
+
gem.version = File.read("VERSION").strip
|
7
|
+
gem.authors = ["Ertugrul Yilmaz"]
|
8
|
+
gem.email = ["***.***.***@gmail.com"]
|
9
|
+
gem.description = %q{Amazon Redshift output plugin for Fluentd}
|
10
|
+
gem.summary = gem.description
|
11
|
+
gem.homepage = "https://github.com/ertugrulyilmaz/fluent-plugin-redshift-out"
|
12
|
+
gem.has_rdoc = false
|
13
|
+
|
14
|
+
gem.files = `git ls-files`.split($/)
|
15
|
+
gem.executables = gem.files.grep(%r{^bin/}).map{ |f| File.basename(f) }
|
16
|
+
gem.test_files = gem.files.grep(%r{^(test|spec|features)/})
|
17
|
+
gem.require_paths = ["lib"]
|
18
|
+
|
19
|
+
gem.add_dependency "fluentd", [">= 0.10.0", "< 2"]
|
20
|
+
gem.add_dependency "aws-sdk-v1", ">= 1.6.3"
|
21
|
+
gem.add_dependency "multi_json", "~> 1.10"
|
22
|
+
gem.add_dependency "yajl-ruby", "~> 1.2"
|
23
|
+
gem.add_dependency "pg", "~> 0.17.0"
|
24
|
+
gem.add_development_dependency "rake"
|
25
|
+
gem.add_development_dependency "simplecov", ">= 0.5.4"
|
26
|
+
gem.add_development_dependency "flexmock", ">= 1.3.1"
|
27
|
+
end
|
@@ -0,0 +1,397 @@
|
|
1
|
+
module Fluent
|
2
|
+
|
3
|
+
|
4
|
+
class RedshiftOutput < BufferedOutput
|
5
|
+
Fluent::Plugin.register_output('redshift-out', self)
|
6
|
+
|
7
|
+
NULL_CHAR_FOR_COPY = "\\N"
|
8
|
+
|
9
|
+
# ignore load table error. (invalid data format)
|
10
|
+
IGNORE_REDSHIFT_ERROR_REGEXP = /^ERROR: Load into table '[^']+' failed\./
|
11
|
+
|
12
|
+
def initialize
|
13
|
+
super
|
14
|
+
require 'aws-sdk-v1'
|
15
|
+
require 'zlib'
|
16
|
+
require 'time'
|
17
|
+
require 'tempfile'
|
18
|
+
require 'pg'
|
19
|
+
require 'csv'
|
20
|
+
require 'multi_json'
|
21
|
+
require 'yajl'
|
22
|
+
::MultiJson.use(:yajl)
|
23
|
+
end
|
24
|
+
|
25
|
+
config_param :record_log_tag, :string, :default => 'log'
|
26
|
+
# s3
|
27
|
+
config_param :aws_key_id, :string, :secret => true, :default => nil
|
28
|
+
config_param :aws_sec_key, :string, :secret => true, :default => nil
|
29
|
+
config_param :s3_bucket, :string
|
30
|
+
config_param :s3_region, :string, :default => nil
|
31
|
+
config_param :path, :string, :default => ""
|
32
|
+
config_param :timestamp_key_format, :string, :default => 'year=%Y/month=%m/day=%d/hour=%H/%Y%m%d-%H%M'
|
33
|
+
config_param :utc, :bool, :default => false
|
34
|
+
# redshift
|
35
|
+
config_param :redshift_host, :string
|
36
|
+
config_param :redshift_port, :integer, :default => 5439
|
37
|
+
config_param :redshift_dbname, :string
|
38
|
+
config_param :redshift_user, :string
|
39
|
+
config_param :redshift_password, :string, :secret => true
|
40
|
+
config_param :redshift_tablename, :string
|
41
|
+
config_param :redshift_schemaname, :string, :default => nil
|
42
|
+
config_param :redshift_copy_base_options, :string , :default => "FILLRECORD ACCEPTANYDATE TRUNCATECOLUMNS"
|
43
|
+
config_param :redshift_copy_options, :string , :default => nil
|
44
|
+
config_param :redshift_connect_timeout, :integer, :default => 10
|
45
|
+
# file format
|
46
|
+
config_param :file_type, :string, :default => nil # json, tsv, csv, msgpack
|
47
|
+
config_param :delimiter, :string, :default => nil
|
48
|
+
# maintenance
|
49
|
+
config_param :maintenance_file_path, :string, :default => nil
|
50
|
+
# for debug
|
51
|
+
config_param :log_suffix, :string, :default => ''
|
52
|
+
|
53
|
+
def configure(conf)
|
54
|
+
super
|
55
|
+
@path = "#{@path}/" unless @path.end_with?('/') # append last slash
|
56
|
+
@path = @path[1..-1] if @path.start_with?('/') # remove head slash
|
57
|
+
@utc = true if conf['utc']
|
58
|
+
@db_conf = {
|
59
|
+
host:@redshift_host,
|
60
|
+
port:@redshift_port,
|
61
|
+
dbname:@redshift_dbname,
|
62
|
+
user:@redshift_user,
|
63
|
+
password:@redshift_password,
|
64
|
+
connect_timeout: @redshift_connect_timeout
|
65
|
+
}
|
66
|
+
@delimiter = determine_delimiter(@file_type) if @delimiter.nil? or @delimiter.empty?
|
67
|
+
$log.debug format_log("redshift file_type:#{@file_type} delimiter:'#{@delimiter}'")
|
68
|
+
@table_name_with_schema = [@redshift_schemaname, @redshift_tablename].compact.join('.')
|
69
|
+
@maintenance_monitor = MaintenanceMonitor.new(@maintenance_file_path)
|
70
|
+
end
|
71
|
+
|
72
|
+
def start
|
73
|
+
super
|
74
|
+
# init s3 conf
|
75
|
+
options = {}
|
76
|
+
if @aws_key_id && @aws_sec_key
|
77
|
+
options[:access_key_id] = @aws_key_id
|
78
|
+
options[:secret_access_key] = @aws_sec_key
|
79
|
+
end
|
80
|
+
options[:region] = @s3_region if @s3_region
|
81
|
+
@s3 = AWS::S3.new(options)
|
82
|
+
@bucket = @s3.buckets[@s3_bucket]
|
83
|
+
@redshift_connection = RedshiftConnection.new(@db_conf)
|
84
|
+
@redshift_connection.connect_start
|
85
|
+
end
|
86
|
+
|
87
|
+
def format(tag, time, record)
|
88
|
+
if json?
|
89
|
+
record.to_msgpack
|
90
|
+
elsif msgpack?
|
91
|
+
{ @record_log_tag => record }.to_msgpack
|
92
|
+
else
|
93
|
+
"#{record[@record_log_tag]}\n"
|
94
|
+
end
|
95
|
+
end
|
96
|
+
|
97
|
+
def write(chunk)
|
98
|
+
$log.debug format_log("start creating gz.")
|
99
|
+
@maintenance_monitor.check_maintenance!
|
100
|
+
|
101
|
+
# create a gz file
|
102
|
+
tmp = Tempfile.new("s3-")
|
103
|
+
tmp =
|
104
|
+
if json? || msgpack?
|
105
|
+
create_gz_file_from_structured_data(tmp, chunk, @delimiter)
|
106
|
+
else
|
107
|
+
create_gz_file_from_flat_data(tmp, chunk)
|
108
|
+
end
|
109
|
+
|
110
|
+
# no data -> skip
|
111
|
+
unless tmp
|
112
|
+
$log.debug format_log("received no valid data. ")
|
113
|
+
return false # for debug
|
114
|
+
end
|
115
|
+
|
116
|
+
# create a file path with time format
|
117
|
+
s3path = create_s3path(@bucket, @path)
|
118
|
+
|
119
|
+
# upload gz to s3
|
120
|
+
@bucket.objects[s3path].write(Pathname.new(tmp.path),
|
121
|
+
:acl => :bucket_owner_full_control)
|
122
|
+
|
123
|
+
# close temp file
|
124
|
+
tmp.close!
|
125
|
+
|
126
|
+
# copy gz on s3 to redshift
|
127
|
+
s3_uri = "s3://#{@s3_bucket}/#{s3path}"
|
128
|
+
credentials = @s3.client.credential_provider.credentials
|
129
|
+
sql = "copy #{@table_name_with_schema} from '#{s3_uri}'"
|
130
|
+
sql += " CREDENTIALS 'aws_access_key_id=#{credentials[:access_key_id]};aws_secret_access_key=#{credentials[:secret_access_key]}"
|
131
|
+
sql += ";token=#{credentials[:session_token]}" if credentials[:session_token]
|
132
|
+
sql += "' delimiter '#{@delimiter}' GZIP ESCAPE #{@redshift_copy_base_options} #{@redshift_copy_options};"
|
133
|
+
|
134
|
+
$log.debug format_log("start copying. s3_uri=#{s3_uri}")
|
135
|
+
|
136
|
+
begin
|
137
|
+
@redshift_connection.exec(sql)
|
138
|
+
$log.info format_log("completed copying to redshift. s3_uri=#{s3_uri}")
|
139
|
+
rescue RedshiftError => e
|
140
|
+
if e.to_s =~ IGNORE_REDSHIFT_ERROR_REGEXP
|
141
|
+
$log.error format_log("failed to copy data into redshift due to load error. s3_uri=#{s3_uri}"), :error=>e.to_s
|
142
|
+
return false # for debug
|
143
|
+
end
|
144
|
+
raise e
|
145
|
+
end
|
146
|
+
true # for debug
|
147
|
+
end
|
148
|
+
|
149
|
+
protected
|
150
|
+
|
151
|
+
def format_log(message)
|
152
|
+
(@log_suffix and not @log_suffix.empty?) ? "#{message} #{@log_suffix}" : message
|
153
|
+
end
|
154
|
+
|
155
|
+
private
|
156
|
+
|
157
|
+
def json?
|
158
|
+
@file_type == 'json'
|
159
|
+
end
|
160
|
+
|
161
|
+
def msgpack?
|
162
|
+
@file_type == 'msgpack'
|
163
|
+
end
|
164
|
+
|
165
|
+
def create_gz_file_from_flat_data(dst_file, chunk)
|
166
|
+
gzw = nil
|
167
|
+
begin
|
168
|
+
gzw = Zlib::GzipWriter.new(dst_file)
|
169
|
+
chunk.write_to(gzw)
|
170
|
+
ensure
|
171
|
+
gzw.close rescue nil if gzw
|
172
|
+
end
|
173
|
+
dst_file
|
174
|
+
end
|
175
|
+
|
176
|
+
def create_gz_file_from_structured_data(dst_file, chunk, delimiter)
|
177
|
+
# fetch the table definition from redshift
|
178
|
+
redshift_table_columns = @redshift_connection.fetch_table_columns(@redshift_tablename, @redshift_schemaname)
|
179
|
+
if redshift_table_columns == nil
|
180
|
+
raise "failed to fetch the redshift table definition."
|
181
|
+
elsif redshift_table_columns.empty?
|
182
|
+
$log.warn format_log("no table on redshift. table_name=#{@table_name_with_schema}")
|
183
|
+
return nil
|
184
|
+
end
|
185
|
+
|
186
|
+
# convert json to tsv format text
|
187
|
+
gzw = nil
|
188
|
+
begin
|
189
|
+
gzw = Zlib::GzipWriter.new(dst_file)
|
190
|
+
chunk.msgpack_each do |record|
|
191
|
+
next unless record
|
192
|
+
begin
|
193
|
+
hash = json? ? json_to_hash(record[@record_log_tag]) : record[@record_log_tag]
|
194
|
+
tsv_text = hash_to_table_text(redshift_table_columns, hash, delimiter)
|
195
|
+
gzw.write(tsv_text) if tsv_text and not tsv_text.empty?
|
196
|
+
rescue => e
|
197
|
+
$log.error format_log("failed to create table text from #{@file_type}. text=(#{record[@record_log_tag]})"), :error=>e.to_s
|
198
|
+
$log.error_backtrace
|
199
|
+
end
|
200
|
+
end
|
201
|
+
return nil unless gzw.pos > 0
|
202
|
+
ensure
|
203
|
+
gzw.close rescue nil if gzw
|
204
|
+
end
|
205
|
+
dst_file
|
206
|
+
end
|
207
|
+
|
208
|
+
def determine_delimiter(file_type)
|
209
|
+
case file_type
|
210
|
+
when 'json', 'msgpack', 'tsv'
|
211
|
+
"\t"
|
212
|
+
when "csv"
|
213
|
+
','
|
214
|
+
else
|
215
|
+
raise Fluent::ConfigError, "Invalid file_type:#{file_type}."
|
216
|
+
end
|
217
|
+
end
|
218
|
+
|
219
|
+
def json_to_hash(json_text)
|
220
|
+
return nil if json_text.to_s.empty?
|
221
|
+
|
222
|
+
MultiJson.load(json_text)
|
223
|
+
rescue => e
|
224
|
+
$log.warn format_log("failed to parse json. "), :error => e.to_s
|
225
|
+
nil
|
226
|
+
end
|
227
|
+
|
228
|
+
def hash_to_table_text(redshift_table_columns, hash, delimiter)
|
229
|
+
return "" unless hash
|
230
|
+
|
231
|
+
# extract values from hash
|
232
|
+
val_list = redshift_table_columns.collect {|cn| hash[cn]}
|
233
|
+
|
234
|
+
if val_list.all?{|v| v.nil?}
|
235
|
+
$log.warn format_log("no data match for table columns on redshift. data=#{hash} table_columns=#{redshift_table_columns}")
|
236
|
+
return ""
|
237
|
+
end
|
238
|
+
|
239
|
+
generate_line_with_delimiter(val_list, delimiter)
|
240
|
+
end
|
241
|
+
|
242
|
+
def generate_line_with_delimiter(val_list, delimiter)
|
243
|
+
val_list.collect do |val|
|
244
|
+
case val
|
245
|
+
when nil
|
246
|
+
NULL_CHAR_FOR_COPY
|
247
|
+
when ''
|
248
|
+
''
|
249
|
+
when Hash, Array
|
250
|
+
escape_text_for_copy(MultiJson.dump(val))
|
251
|
+
else
|
252
|
+
escape_text_for_copy(val.to_s)
|
253
|
+
end
|
254
|
+
end.join(delimiter) + "\n"
|
255
|
+
end
|
256
|
+
|
257
|
+
def escape_text_for_copy(val)
|
258
|
+
val.gsub(/\\|\t|\n/, {"\\" => "\\\\", "\t" => "\\\t", "\n" => "\\\n"}) # escape tab, newline and backslash
|
259
|
+
end
|
260
|
+
|
261
|
+
def create_s3path(bucket, path)
|
262
|
+
timestamp_key = (@utc) ? Time.now.utc.strftime(@timestamp_key_format) : Time.now.strftime(@timestamp_key_format)
|
263
|
+
i = 0
|
264
|
+
begin
|
265
|
+
suffix = "_#{'%02d' % i}"
|
266
|
+
s3path = "#{path}#{timestamp_key}#{suffix}.gz"
|
267
|
+
i += 1
|
268
|
+
end while bucket.objects[s3path].exists?
|
269
|
+
s3path
|
270
|
+
end
|
271
|
+
|
272
|
+
class RedshiftError < StandardError
|
273
|
+
def initialize(msg)
|
274
|
+
case msg
|
275
|
+
when PG::Error
|
276
|
+
@pg_error = msg
|
277
|
+
super(msg.to_s)
|
278
|
+
set_backtrace(msg.backtrace)
|
279
|
+
else
|
280
|
+
super
|
281
|
+
end
|
282
|
+
end
|
283
|
+
|
284
|
+
attr_accessor :pg_error
|
285
|
+
end
|
286
|
+
|
287
|
+
class RedshiftConnection
|
288
|
+
REDSHIFT_CONNECT_TIMEOUT = 10.0 # 10sec
|
289
|
+
|
290
|
+
def initialize(db_conf)
|
291
|
+
@db_conf = db_conf
|
292
|
+
@connection = nil
|
293
|
+
ObjectSpace.define_finalizer(self) {
|
294
|
+
close()
|
295
|
+
}
|
296
|
+
end
|
297
|
+
|
298
|
+
attr_reader :db_conf
|
299
|
+
|
300
|
+
def fetch_table_columns(table_name, schema_name)
|
301
|
+
columns = nil
|
302
|
+
exec(fetch_columns_sql(table_name, schema_name)) do |result|
|
303
|
+
columns = result.collect{|row| row['column_name']}
|
304
|
+
end
|
305
|
+
columns
|
306
|
+
end
|
307
|
+
|
308
|
+
def exec(sql, &block)
|
309
|
+
conn = @connection
|
310
|
+
conn = create_redshift_connection if conn.nil?
|
311
|
+
if block
|
312
|
+
conn.exec(sql) {|result| block.call(result)}
|
313
|
+
else
|
314
|
+
conn.exec(sql)
|
315
|
+
end
|
316
|
+
rescue PG::Error => e
|
317
|
+
raise RedshiftError.new(e)
|
318
|
+
ensure
|
319
|
+
conn.close if conn && @connection.nil?
|
320
|
+
end
|
321
|
+
|
322
|
+
def connect_start
|
323
|
+
@connection = create_redshift_connection
|
324
|
+
end
|
325
|
+
|
326
|
+
def close
|
327
|
+
@connection.close rescue nil if @connection
|
328
|
+
@connection = nil
|
329
|
+
end
|
330
|
+
|
331
|
+
private
|
332
|
+
|
333
|
+
def create_redshift_connection
|
334
|
+
hostaddr = IPSocket.getaddress(db_conf[:host])
|
335
|
+
db_conf[:hostaddr] = hostaddr
|
336
|
+
|
337
|
+
conn = PG::Connection.connect_start(db_conf)
|
338
|
+
raise RedshiftError.new("Unable to create a new connection.") unless conn
|
339
|
+
if conn.status == PG::CONNECTION_BAD
|
340
|
+
raise RedshiftError.new("Connection failed: %s" % [ conn.error_message ])
|
341
|
+
end
|
342
|
+
|
343
|
+
socket = conn.socket_io
|
344
|
+
poll_status = PG::PGRES_POLLING_WRITING
|
345
|
+
until poll_status == PG::PGRES_POLLING_OK || poll_status == PG::PGRES_POLLING_FAILED
|
346
|
+
case poll_status
|
347
|
+
when PG::PGRES_POLLING_READING
|
348
|
+
IO.select([socket], nil, nil, REDSHIFT_CONNECT_TIMEOUT) or
|
349
|
+
raise RedshiftError.new("Asynchronous connection timed out!(READING)")
|
350
|
+
when PG::PGRES_POLLING_WRITING
|
351
|
+
IO.select(nil, [socket], nil, REDSHIFT_CONNECT_TIMEOUT) or
|
352
|
+
raise RedshiftError.new("Asynchronous connection timed out!(WRITING)")
|
353
|
+
end
|
354
|
+
poll_status = conn.connect_poll
|
355
|
+
end
|
356
|
+
|
357
|
+
unless conn.status == PG::CONNECTION_OK
|
358
|
+
raise RedshiftError, ("Connect failed: %s" % [conn.error_message.to_s.lines.uniq.join(" ")])
|
359
|
+
end
|
360
|
+
|
361
|
+
conn
|
362
|
+
rescue => e
|
363
|
+
conn.close rescue nil if conn
|
364
|
+
raise RedshiftError.new(e) if e.kind_of?(PG::Error)
|
365
|
+
raise e
|
366
|
+
end
|
367
|
+
|
368
|
+
def fetch_columns_sql(table_name, schema_name = nil)
|
369
|
+
sql = "select column_name from INFORMATION_SCHEMA.COLUMNS where table_name = '#{table_name}'"
|
370
|
+
sql << " and table_schema = '#{schema_name}'" if schema_name
|
371
|
+
sql << " order by ordinal_position;"
|
372
|
+
sql
|
373
|
+
end
|
374
|
+
end
|
375
|
+
|
376
|
+
class MaintenanceError < StandardError
|
377
|
+
end
|
378
|
+
|
379
|
+
class MaintenanceMonitor
|
380
|
+
def initialize(maintenance_file_path)
|
381
|
+
@file_path = maintenance_file_path
|
382
|
+
end
|
383
|
+
|
384
|
+
def in_maintenance?
|
385
|
+
!!(@file_path && File.exists?(@file_path))
|
386
|
+
end
|
387
|
+
|
388
|
+
def check_maintenance!
|
389
|
+
if in_maintenance?
|
390
|
+
raise MaintenanceError.new("Service is in maintenance mode - maintenance_file_path:#{@file_path}")
|
391
|
+
end
|
392
|
+
end
|
393
|
+
end
|
394
|
+
end
|
395
|
+
|
396
|
+
|
397
|
+
end
|
@@ -0,0 +1,503 @@
|
|
1
|
+
require 'test_helper'
|
2
|
+
|
3
|
+
require 'fluent/test'
|
4
|
+
require 'fluent/plugin/out_redshift-out'
|
5
|
+
require 'flexmock/test_unit'
|
6
|
+
require 'zlib'
|
7
|
+
|
8
|
+
|
9
|
+
class RedshiftOutputTest < Test::Unit::TestCase
|
10
|
+
def setup
|
11
|
+
require 'aws-sdk-v1'
|
12
|
+
require 'pg'
|
13
|
+
require 'csv'
|
14
|
+
Fluent::Test.setup
|
15
|
+
PG::Error.module_eval { attr_accessor :result}
|
16
|
+
end
|
17
|
+
|
18
|
+
MAINTENANCE_FILE_PATH_FOR_TEST = "/tmp/fluentd_redshift_plugin_test_maintenance"
|
19
|
+
|
20
|
+
CONFIG_BASE= %[
|
21
|
+
aws_key_id test_key_id
|
22
|
+
aws_sec_key test_sec_key
|
23
|
+
s3_bucket test_bucket
|
24
|
+
path log
|
25
|
+
redshift_host test_host
|
26
|
+
redshift_dbname test_db
|
27
|
+
redshift_user test_user
|
28
|
+
redshift_password test_password
|
29
|
+
redshift_tablename test_table
|
30
|
+
buffer_type memory
|
31
|
+
utc
|
32
|
+
log_suffix id:5 host:localhost
|
33
|
+
maintenance_file_path #{MAINTENANCE_FILE_PATH_FOR_TEST}
|
34
|
+
]
|
35
|
+
CONFIG_CSV= %[
|
36
|
+
#{CONFIG_BASE}
|
37
|
+
file_type csv
|
38
|
+
]
|
39
|
+
CONFIG_TSV= %[
|
40
|
+
#{CONFIG_BASE}
|
41
|
+
file_type tsv
|
42
|
+
]
|
43
|
+
CONFIG_JSON = %[
|
44
|
+
#{CONFIG_BASE}
|
45
|
+
file_type json
|
46
|
+
]
|
47
|
+
CONFIG_JSON_WITH_SCHEMA = %[
|
48
|
+
#{CONFIG_BASE}
|
49
|
+
redshift_schemaname test_schema
|
50
|
+
file_type json
|
51
|
+
]
|
52
|
+
CONFIG_MSGPACK = %[
|
53
|
+
#{CONFIG_BASE}
|
54
|
+
file_type msgpack
|
55
|
+
]
|
56
|
+
CONFIG_PIPE_DELIMITER= %[
|
57
|
+
#{CONFIG_BASE}
|
58
|
+
delimiter |
|
59
|
+
]
|
60
|
+
CONFIG_PIPE_DELIMITER_WITH_NAME= %[
|
61
|
+
#{CONFIG_BASE}
|
62
|
+
file_type pipe
|
63
|
+
delimiter |
|
64
|
+
]
|
65
|
+
CONFIG=CONFIG_CSV
|
66
|
+
|
67
|
+
RECORD_CSV_A = {"log" => %[val_a,val_b,val_c,val_d]}
|
68
|
+
RECORD_CSV_B = {"log" => %[val_e,val_f,val_g,val_h]}
|
69
|
+
RECORD_TSV_A = {"log" => %[val_a\tval_b\tval_c\tval_d]}
|
70
|
+
RECORD_TSV_B = {"log" => %[val_e\tval_f\tval_g\tval_h]}
|
71
|
+
RECORD_JSON_A = {"log" => %[{"key_a" : "val_a", "key_b" : "val_b"}]}
|
72
|
+
RECORD_JSON_B = {"log" => %[{"key_c" : "val_c", "key_d" : "val_d"}]}
|
73
|
+
RECORD_MSGPACK_A = {"key_a" => "val_a", "key_b" => "val_b"}
|
74
|
+
RECORD_MSGPACK_B = {"key_c" => "val_c", "key_d" => "val_d"}
|
75
|
+
DEFAULT_TIME = Time.parse("2013-03-06 12:15:02 UTC").to_i
|
76
|
+
|
77
|
+
def create_driver(conf = CONFIG, tag='test.input')
|
78
|
+
Fluent::Test::BufferedOutputTestDriver.new(Fluent::RedshiftOutput, tag).configure(conf)
|
79
|
+
end
|
80
|
+
|
81
|
+
def create_driver_no_write(conf = CONFIG, tag='test.input')
|
82
|
+
Fluent::Test::BufferedOutputTestDriver.new(Fluent::RedshiftOutput, tag) do
|
83
|
+
def write(chunk)
|
84
|
+
chunk.read
|
85
|
+
end
|
86
|
+
end.configure(conf)
|
87
|
+
end
|
88
|
+
|
89
|
+
def test_configure
|
90
|
+
assert_raise(Fluent::ConfigError) {
|
91
|
+
d = create_driver('')
|
92
|
+
}
|
93
|
+
assert_raise(Fluent::ConfigError) {
|
94
|
+
d = create_driver(CONFIG_BASE)
|
95
|
+
}
|
96
|
+
d = create_driver(CONFIG_CSV)
|
97
|
+
assert_equal "test_key_id", d.instance.aws_key_id
|
98
|
+
assert_equal "test_sec_key", d.instance.aws_sec_key
|
99
|
+
assert_equal "test_bucket", d.instance.s3_bucket
|
100
|
+
assert_equal "log/", d.instance.path
|
101
|
+
assert_equal "test_host", d.instance.redshift_host
|
102
|
+
assert_equal 5439, d.instance.redshift_port
|
103
|
+
assert_equal "test_db", d.instance.redshift_dbname
|
104
|
+
assert_equal "test_user", d.instance.redshift_user
|
105
|
+
assert_equal "test_password", d.instance.redshift_password
|
106
|
+
assert_equal "test_table", d.instance.redshift_tablename
|
107
|
+
assert_equal nil, d.instance.redshift_schemaname
|
108
|
+
assert_equal "FILLRECORD ACCEPTANYDATE TRUNCATECOLUMNS", d.instance.redshift_copy_base_options
|
109
|
+
assert_equal nil, d.instance.redshift_copy_options
|
110
|
+
assert_equal "csv", d.instance.file_type
|
111
|
+
assert_equal ",", d.instance.delimiter
|
112
|
+
assert_equal true, d.instance.utc
|
113
|
+
assert_equal MAINTENANCE_FILE_PATH_FOR_TEST, d.instance.maintenance_file_path
|
114
|
+
end
|
115
|
+
def test_configure_with_schemaname
|
116
|
+
d = create_driver(CONFIG_JSON_WITH_SCHEMA)
|
117
|
+
assert_equal "test_schema", d.instance.redshift_schemaname
|
118
|
+
end
|
119
|
+
def test_configure_localtime
|
120
|
+
d = create_driver(CONFIG_CSV.gsub(/ *utc */, ''))
|
121
|
+
assert_equal false, d.instance.utc
|
122
|
+
end
|
123
|
+
def test_configure_no_path
|
124
|
+
d = create_driver(CONFIG_CSV.gsub(/ *path *.+$/, ''))
|
125
|
+
assert_equal "", d.instance.path
|
126
|
+
end
|
127
|
+
def test_configure_root_path
|
128
|
+
d = create_driver(CONFIG_CSV.gsub(/ *path *.+$/, 'path /'))
|
129
|
+
assert_equal "", d.instance.path
|
130
|
+
end
|
131
|
+
def test_configure_path_with_slash
|
132
|
+
d = create_driver(CONFIG_CSV.gsub(/ *path *.+$/, 'path log/'))
|
133
|
+
assert_equal "log/", d.instance.path
|
134
|
+
end
|
135
|
+
def test_configure_path_starts_with_slash
|
136
|
+
d = create_driver(CONFIG_CSV.gsub(/ *path *.+$/, 'path /log/'))
|
137
|
+
assert_equal "log/", d.instance.path
|
138
|
+
end
|
139
|
+
def test_configure_path_starts_with_slash_without_last_slash
|
140
|
+
d = create_driver(CONFIG_CSV.gsub(/ *path *.+$/, 'path /log'))
|
141
|
+
assert_equal "log/", d.instance.path
|
142
|
+
end
|
143
|
+
def test_configure_tsv
|
144
|
+
d1 = create_driver(CONFIG_TSV)
|
145
|
+
assert_equal "tsv", d1.instance.file_type
|
146
|
+
assert_equal "\t", d1.instance.delimiter
|
147
|
+
end
|
148
|
+
def test_configure_json
|
149
|
+
d2 = create_driver(CONFIG_JSON)
|
150
|
+
assert_equal "json", d2.instance.file_type
|
151
|
+
assert_equal "\t", d2.instance.delimiter
|
152
|
+
end
|
153
|
+
def test_configure_msgpack
|
154
|
+
d2 = create_driver(CONFIG_MSGPACK)
|
155
|
+
assert_equal "msgpack", d2.instance.file_type
|
156
|
+
assert_equal "\t", d2.instance.delimiter
|
157
|
+
end
|
158
|
+
def test_configure_original_file_type
|
159
|
+
d3 = create_driver(CONFIG_PIPE_DELIMITER)
|
160
|
+
assert_equal nil, d3.instance.file_type
|
161
|
+
assert_equal "|", d3.instance.delimiter
|
162
|
+
|
163
|
+
d4 = create_driver(CONFIG_PIPE_DELIMITER_WITH_NAME)
|
164
|
+
assert_equal "pipe", d4.instance.file_type
|
165
|
+
assert_equal "|", d4.instance.delimiter
|
166
|
+
end
|
167
|
+
def test_configure_no_log_suffix
|
168
|
+
d = create_driver(CONFIG_CSV.gsub(/ *log_suffix *.+$/, ''))
|
169
|
+
assert_equal "", d.instance.log_suffix
|
170
|
+
end
|
171
|
+
|
172
|
+
def emit_csv(d)
|
173
|
+
d.emit(RECORD_CSV_A, DEFAULT_TIME)
|
174
|
+
d.emit(RECORD_CSV_B, DEFAULT_TIME)
|
175
|
+
end
|
176
|
+
def emit_tsv(d)
|
177
|
+
d.emit(RECORD_TSV_A, DEFAULT_TIME)
|
178
|
+
d.emit(RECORD_TSV_B, DEFAULT_TIME)
|
179
|
+
end
|
180
|
+
def emit_json(d)
|
181
|
+
d.emit(RECORD_JSON_A, DEFAULT_TIME)
|
182
|
+
d.emit(RECORD_JSON_B, DEFAULT_TIME)
|
183
|
+
end
|
184
|
+
def emit_msgpack(d)
|
185
|
+
d.emit(RECORD_MSGPACK_A, DEFAULT_TIME)
|
186
|
+
d.emit(RECORD_MSGPACK_B, DEFAULT_TIME)
|
187
|
+
end
|
188
|
+
|
189
|
+
def test_format_csv
|
190
|
+
setup_mocks("")
|
191
|
+
d_csv = create_driver_no_write(CONFIG_CSV)
|
192
|
+
emit_csv(d_csv)
|
193
|
+
d_csv.expect_format RECORD_CSV_A['log'] + "\n"
|
194
|
+
d_csv.expect_format RECORD_CSV_B['log'] + "\n"
|
195
|
+
d_csv.run
|
196
|
+
end
|
197
|
+
def test_format_tsv
|
198
|
+
setup_mocks("")
|
199
|
+
d_tsv = create_driver_no_write(CONFIG_TSV)
|
200
|
+
emit_tsv(d_tsv)
|
201
|
+
d_tsv.expect_format RECORD_TSV_A['log'] + "\n"
|
202
|
+
d_tsv.expect_format RECORD_TSV_B['log'] + "\n"
|
203
|
+
d_tsv.run
|
204
|
+
end
|
205
|
+
def test_format_json
|
206
|
+
setup_mocks("")
|
207
|
+
d_json = create_driver_no_write(CONFIG_JSON)
|
208
|
+
emit_json(d_json)
|
209
|
+
d_json.expect_format RECORD_JSON_A.to_msgpack
|
210
|
+
d_json.expect_format RECORD_JSON_B.to_msgpack
|
211
|
+
d_json.run
|
212
|
+
end
|
213
|
+
|
214
|
+
def test_format_msgpack
|
215
|
+
setup_mocks("")
|
216
|
+
d_msgpack = create_driver_no_write(CONFIG_MSGPACK)
|
217
|
+
emit_msgpack(d_msgpack)
|
218
|
+
d_msgpack.expect_format({ 'log' => RECORD_MSGPACK_A }.to_msgpack)
|
219
|
+
d_msgpack.expect_format({ 'log' => RECORD_MSGPACK_B }.to_msgpack)
|
220
|
+
d_msgpack.run
|
221
|
+
end
|
222
|
+
|
223
|
+
def setup_redshift_connection_mock(options = {})
|
224
|
+
options ||= {}
|
225
|
+
column_names = options[:column_names] || ['key_a', 'key_b', 'key_c', 'key_d', 'key_e', 'key_f', 'key_g', 'key_h']
|
226
|
+
schema_name = options[:schema_name]
|
227
|
+
table_name = options[:table_name] || 'test_table'
|
228
|
+
exec_sql_proc = options[:exec_sql_proc]
|
229
|
+
|
230
|
+
column_list_query_regex =
|
231
|
+
if schema_name
|
232
|
+
/\Aselect column_name from INFORMATION_SCHEMA.COLUMNS where table_name = '#{table_name}' and table_schema = '#{schema_name}'/
|
233
|
+
else
|
234
|
+
/\Aselect column_name from INFORMATION_SCHEMA.COLUMNS where table_name = '#{table_name}'/
|
235
|
+
end
|
236
|
+
copy_query_regex =
|
237
|
+
if schema_name
|
238
|
+
/\Acopy #{schema_name}.#{table_name} from/
|
239
|
+
else
|
240
|
+
/\Acopy #{table_name} from/
|
241
|
+
end
|
242
|
+
|
243
|
+
flexmock(Fluent::RedshiftOutput::RedshiftConnection).new_instances do |conn|
|
244
|
+
conn.should_receive(:exec).and_return do |sql, block|
|
245
|
+
if exec_sql_proc
|
246
|
+
exec_sql_proc.call(sql, block)
|
247
|
+
elsif block
|
248
|
+
if sql =~ column_list_query_regex
|
249
|
+
block.call column_names.collect{|key| {'column_name' => key}}
|
250
|
+
else
|
251
|
+
block.call []
|
252
|
+
end
|
253
|
+
else
|
254
|
+
unless sql =~ copy_query_regex
|
255
|
+
error = PG::Error.new("ERROR: Load into table '#{@target_table}' failed. Check 'stl_load_errors' system table for details.")
|
256
|
+
error.result = "ERROR: Load into table '#{@target_table}' failed. Check 'stl_load_errors' system table for details."
|
257
|
+
raise Fluent::RedshiftOutput::RedshiftError.new(error)
|
258
|
+
end
|
259
|
+
end
|
260
|
+
end
|
261
|
+
conn.should_receive(:connect_start)
|
262
|
+
end
|
263
|
+
end
|
264
|
+
|
265
|
+
def setup_s3_mock(expected_data)
|
266
|
+
current_time = Time.now
|
267
|
+
|
268
|
+
# create mock of s3 object
|
269
|
+
s3obj = flexmock(AWS::S3::S3Object)
|
270
|
+
s3obj.should_receive(:exists?).with_any_args.and_return { false }
|
271
|
+
s3obj.should_receive(:write).with(
|
272
|
+
# pathname
|
273
|
+
on { |pathname|
|
274
|
+
data = nil
|
275
|
+
pathname.open { |f|
|
276
|
+
gz = Zlib::GzipReader.new(f)
|
277
|
+
data = gz.read
|
278
|
+
gz.close
|
279
|
+
}
|
280
|
+
assert_equal expected_data, data
|
281
|
+
},
|
282
|
+
:acl => :bucket_owner_full_control
|
283
|
+
).and_return { true }
|
284
|
+
|
285
|
+
# create mock of s3 object collection
|
286
|
+
s3obj_col = flexmock(AWS::S3::ObjectCollection)
|
287
|
+
s3obj_col.should_receive(:[]).with(
|
288
|
+
on { |key|
|
289
|
+
expected_key = current_time.utc.strftime("log/year=%Y/month=%m/day=%d/hour=%H/%Y%m%d-%H%M_00.gz")
|
290
|
+
key == expected_key
|
291
|
+
}).
|
292
|
+
and_return {
|
293
|
+
s3obj
|
294
|
+
}
|
295
|
+
|
296
|
+
# create mock of s3 bucket
|
297
|
+
flexmock(AWS::S3::Bucket).new_instances do |bucket|
|
298
|
+
bucket.should_receive(:objects).with_any_args.
|
299
|
+
and_return {
|
300
|
+
s3obj_col
|
301
|
+
}
|
302
|
+
end
|
303
|
+
end
|
304
|
+
|
305
|
+
def setup_tempfile_mock_to_be_closed
|
306
|
+
flexmock(Tempfile).new_instances.should_receive(:close!).at_least.once
|
307
|
+
end
|
308
|
+
|
309
|
+
def setup_mocks(expected_data, options = {})
|
310
|
+
setup_redshift_connection_mock(options)
|
311
|
+
setup_s3_mock(expected_data)
|
312
|
+
end
|
313
|
+
|
314
|
+
def test_write_with_csv
|
315
|
+
setup_mocks(%[val_a,val_b,val_c,val_d\nval_e,val_f,val_g,val_h\n])
|
316
|
+
setup_tempfile_mock_to_be_closed
|
317
|
+
d_csv = create_driver
|
318
|
+
emit_csv(d_csv)
|
319
|
+
assert_equal true, d_csv.run
|
320
|
+
end
|
321
|
+
|
322
|
+
def test_write_with_json
|
323
|
+
setup_mocks(%[val_a\tval_b\t\\N\t\\N\t\\N\t\\N\t\\N\t\\N\n\\N\t\\N\tval_c\tval_d\t\\N\t\\N\t\\N\t\\N\n])
|
324
|
+
setup_tempfile_mock_to_be_closed
|
325
|
+
d_json = create_driver(CONFIG_JSON)
|
326
|
+
emit_json(d_json)
|
327
|
+
assert_equal true, d_json.run
|
328
|
+
end
|
329
|
+
|
330
|
+
def test_write_with_json_hash_value
|
331
|
+
setup_mocks("val_a\t{\"foo\":\"var\"}\t\\N\t\\N\t\\N\t\\N\t\\N\t\\N\n\\N\t\\N\tval_c\tval_d\t\\N\t\\N\t\\N\t\\N\n")
|
332
|
+
d_json = create_driver(CONFIG_JSON)
|
333
|
+
d_json.emit({"log" => %[{"key_a" : "val_a", "key_b" : {"foo" : "var"}}]} , DEFAULT_TIME)
|
334
|
+
d_json.emit(RECORD_JSON_B, DEFAULT_TIME)
|
335
|
+
assert_equal true, d_json.run
|
336
|
+
end
|
337
|
+
|
338
|
+
def test_write_with_json_array_value
|
339
|
+
setup_mocks("val_a\t[\"foo\",\"var\"]\t\\N\t\\N\t\\N\t\\N\t\\N\t\\N\n\\N\t\\N\tval_c\tval_d\t\\N\t\\N\t\\N\t\\N\n")
|
340
|
+
d_json = create_driver(CONFIG_JSON)
|
341
|
+
d_json.emit({"log" => %[{"key_a" : "val_a", "key_b" : ["foo", "var"]}]} , DEFAULT_TIME)
|
342
|
+
d_json.emit(RECORD_JSON_B, DEFAULT_TIME)
|
343
|
+
assert_equal true, d_json.run
|
344
|
+
end
|
345
|
+
|
346
|
+
def test_write_with_json_including_tab_newline_quote
|
347
|
+
setup_mocks("val_a_with_\\\t_tab_\\\n_newline\tval_b_with_\\\\_quote\t\\N\t\\N\t\\N\t\\N\t\\N\t\\N\n\\N\t\\N\tval_c\tval_d\t\\N\t\\N\t\\N\t\\N\n")
|
348
|
+
d_json = create_driver(CONFIG_JSON)
|
349
|
+
d_json.emit({"log" => %[{"key_a" : "val_a_with_\\t_tab_\\n_newline", "key_b" : "val_b_with_\\\\_quote"}]} , DEFAULT_TIME)
|
350
|
+
d_json.emit(RECORD_JSON_B, DEFAULT_TIME)
|
351
|
+
assert_equal true, d_json.run
|
352
|
+
end
|
353
|
+
|
354
|
+
def test_write_with_json_empty_text_value
|
355
|
+
setup_mocks(%[val_a\t\t\\N\t\\N\t\\N\t\\N\t\\N\t\\N\n])
|
356
|
+
d_json = create_driver(CONFIG_JSON)
|
357
|
+
d_json.emit({"log" => %[{"key_a" : "val_a", "key_b" : ""}]} , DEFAULT_TIME)
|
358
|
+
assert_equal true, d_json.run
|
359
|
+
end
|
360
|
+
|
361
|
+
def test_write_with_json_no_data
|
362
|
+
setup_mocks("")
|
363
|
+
d_json = create_driver(CONFIG_JSON)
|
364
|
+
d_json.emit("", DEFAULT_TIME)
|
365
|
+
d_json.emit("", DEFAULT_TIME)
|
366
|
+
assert_equal false, d_json.run
|
367
|
+
end
|
368
|
+
|
369
|
+
def test_write_with_json_invalid_one_line
|
370
|
+
setup_mocks(%[\\N\t\\N\tval_c\tval_d\t\\N\t\\N\t\\N\t\\N\n])
|
371
|
+
d_json = create_driver(CONFIG_JSON)
|
372
|
+
d_json.emit({"log" => %[}}]}, DEFAULT_TIME)
|
373
|
+
d_json.emit(RECORD_JSON_B, DEFAULT_TIME)
|
374
|
+
assert_equal true, d_json.run
|
375
|
+
end
|
376
|
+
|
377
|
+
def test_write_with_json_no_available_data
|
378
|
+
setup_mocks(%[val_a\tval_b\t\\N\t\\N\t\\N\t\\N\t\\N\t\\N\n])
|
379
|
+
d_json = create_driver(CONFIG_JSON)
|
380
|
+
d_json.emit(RECORD_JSON_A, DEFAULT_TIME)
|
381
|
+
d_json.emit({"log" => %[{"key_o" : "val_o", "key_p" : "val_p"}]}, DEFAULT_TIME)
|
382
|
+
assert_equal true, d_json.run
|
383
|
+
end
|
384
|
+
|
385
|
+
def test_write_with_msgpack
|
386
|
+
setup_mocks(%[val_a\tval_b\t\\N\t\\N\t\\N\t\\N\t\\N\t\\N\n\\N\t\\N\tval_c\tval_d\t\\N\t\\N\t\\N\t\\N\n])
|
387
|
+
d_msgpack = create_driver(CONFIG_MSGPACK)
|
388
|
+
emit_msgpack(d_msgpack)
|
389
|
+
assert_equal true, d_msgpack.run
|
390
|
+
end
|
391
|
+
|
392
|
+
def test_write_with_msgpack_hash_value
|
393
|
+
setup_mocks("val_a\t{\"foo\":\"var\"}\t\\N\t\\N\t\\N\t\\N\t\\N\t\\N\n\\N\t\\N\tval_c\tval_d\t\\N\t\\N\t\\N\t\\N\n")
|
394
|
+
d_msgpack = create_driver(CONFIG_MSGPACK)
|
395
|
+
d_msgpack.emit({"key_a" => "val_a", "key_b" => {"foo" => "var"}} , DEFAULT_TIME)
|
396
|
+
d_msgpack.emit(RECORD_MSGPACK_B, DEFAULT_TIME)
|
397
|
+
assert_equal true, d_msgpack.run
|
398
|
+
end
|
399
|
+
|
400
|
+
def test_write_with_msgpack_array_value
|
401
|
+
setup_mocks("val_a\t[\"foo\",\"var\"]\t\\N\t\\N\t\\N\t\\N\t\\N\t\\N\n\\N\t\\N\tval_c\tval_d\t\\N\t\\N\t\\N\t\\N\n")
|
402
|
+
d_msgpack = create_driver(CONFIG_MSGPACK)
|
403
|
+
d_msgpack.emit({"key_a" => "val_a", "key_b" => ["foo", "var"]} , DEFAULT_TIME)
|
404
|
+
d_msgpack.emit(RECORD_MSGPACK_B, DEFAULT_TIME)
|
405
|
+
assert_equal true, d_msgpack.run
|
406
|
+
end
|
407
|
+
|
408
|
+
def test_write_with_msgpack_including_tab_newline_quote
|
409
|
+
setup_mocks("val_a_with_\\\t_tab_\\\n_newline\tval_b_with_\\\\_quote\t\\N\t\\N\t\\N\t\\N\t\\N\t\\N\n\\N\t\\N\tval_c\tval_d\t\\N\t\\N\t\\N\t\\N\n")
|
410
|
+
d_msgpack = create_driver(CONFIG_MSGPACK)
|
411
|
+
d_msgpack.emit({"key_a" => "val_a_with_\t_tab_\n_newline", "key_b" => "val_b_with_\\_quote"} , DEFAULT_TIME)
|
412
|
+
d_msgpack.emit(RECORD_MSGPACK_B, DEFAULT_TIME)
|
413
|
+
assert_equal true, d_msgpack.run
|
414
|
+
end
|
415
|
+
|
416
|
+
def test_write_with_msgpack_no_data
|
417
|
+
setup_mocks("")
|
418
|
+
d_msgpack = create_driver(CONFIG_MSGPACK)
|
419
|
+
d_msgpack.emit({}, DEFAULT_TIME)
|
420
|
+
d_msgpack.emit({}, DEFAULT_TIME)
|
421
|
+
assert_equal false, d_msgpack.run
|
422
|
+
end
|
423
|
+
|
424
|
+
def test_write_with_msgpack_no_available_data
|
425
|
+
setup_mocks(%[val_a\tval_b\t\\N\t\\N\t\\N\t\\N\t\\N\t\\N\n])
|
426
|
+
d_msgpack = create_driver(CONFIG_MSGPACK)
|
427
|
+
d_msgpack.emit(RECORD_MSGPACK_A, DEFAULT_TIME)
|
428
|
+
d_msgpack.emit({"key_o" => "val_o", "key_p" => "val_p"}, DEFAULT_TIME)
|
429
|
+
assert_equal true, d_msgpack.run
|
430
|
+
end
|
431
|
+
|
432
|
+
def test_write_redshift_connection_error
|
433
|
+
setup_mocks(%[val_a,val_b,val_c,val_d\nval_e,val_f,val_g,val_h\n],
|
434
|
+
exec_sql_proc: Proc.new {|sql, block| raise Fluent::RedshiftOutput::RedshiftError, "redshift connection error" })
|
435
|
+
d_csv = create_driver
|
436
|
+
emit_csv(d_csv)
|
437
|
+
assert_raise(Fluent::RedshiftOutput::RedshiftError) {
|
438
|
+
d_csv.run
|
439
|
+
}
|
440
|
+
end
|
441
|
+
|
442
|
+
def test_write_redshift_load_error
|
443
|
+
setup_mocks(%[val_a,val_b,val_c,val_d\nval_e,val_f,val_g,val_h\n],
|
444
|
+
exec_sql_proc: Proc.new {|sql, block|
|
445
|
+
msg = "ERROR: Load into table 'apache_log' failed. Check 'stl_load_errors' system table for details."
|
446
|
+
raise Fluent::RedshiftOutput::RedshiftError.new(msg)
|
447
|
+
})
|
448
|
+
|
449
|
+
d_csv = create_driver
|
450
|
+
emit_csv(d_csv)
|
451
|
+
assert_equal false, d_csv.run
|
452
|
+
end
|
453
|
+
|
454
|
+
def test_write_with_json_redshift_connection_error
|
455
|
+
setup_mocks(%[val_a,val_b,val_c,val_d\nval_e,val_f,val_g,val_h\n],
|
456
|
+
exec_sql_proc: Proc.new {|sql, block| raise Fluent::RedshiftOutput::RedshiftError.new("redshift connection error")})
|
457
|
+
|
458
|
+
d_json = create_driver(CONFIG_JSON)
|
459
|
+
emit_json(d_json)
|
460
|
+
assert_raise(Fluent::RedshiftOutput::RedshiftError) {
|
461
|
+
d_json.run
|
462
|
+
}
|
463
|
+
end
|
464
|
+
|
465
|
+
def test_write_with_json_no_table_on_redshift
|
466
|
+
setup_mocks(%[val_a,val_b,val_c,val_d\nval_e,val_f,val_g,val_h\n],
|
467
|
+
exec_sql_proc: Proc.new {|sql, block| block.call [] if block })
|
468
|
+
|
469
|
+
d_json = create_driver(CONFIG_JSON)
|
470
|
+
emit_json(d_json)
|
471
|
+
assert_equal false, d_json.run
|
472
|
+
end
|
473
|
+
|
474
|
+
def test_write_with_json_failed_to_get_columns
|
475
|
+
setup_mocks("", exec_sql_proc: Proc.new {|sql, block| nil})
|
476
|
+
|
477
|
+
d_json = create_driver(CONFIG_JSON)
|
478
|
+
emit_json(d_json)
|
479
|
+
assert_raise(RuntimeError, "failed to fetch the redshift table definition.") {
|
480
|
+
d_json.run
|
481
|
+
}
|
482
|
+
end
|
483
|
+
|
484
|
+
def test_write_with_json_fetch_column_with_schema
|
485
|
+
setup_mocks(%[val_a\tval_b\t\\N\t\\N\t\\N\t\\N\t\\N\t\\N\n\\N\t\\N\tval_c\tval_d\t\\N\t\\N\t\\N\t\\N\n],
|
486
|
+
schema_name: 'test_schema')
|
487
|
+
d_json = create_driver(CONFIG_JSON_WITH_SCHEMA)
|
488
|
+
emit_json(d_json)
|
489
|
+
assert_equal true, d_json.run
|
490
|
+
end
|
491
|
+
|
492
|
+
def test_maintenance_mode
|
493
|
+
setup_mocks("")
|
494
|
+
flexmock(File).should_receive(:exists?).with(MAINTENANCE_FILE_PATH_FOR_TEST).and_return(true)
|
495
|
+
|
496
|
+
d_json = create_driver(CONFIG_JSON)
|
497
|
+
emit_json(d_json)
|
498
|
+
assert_raise(Fluent::RedshiftOutput::MaintenanceError,
|
499
|
+
"Service is in maintenance mode - maintenance_file_path:#{MAINTENANCE_FILE_PATH_FOR_TEST}") {
|
500
|
+
d_json.run
|
501
|
+
}
|
502
|
+
end
|
503
|
+
end
|
data/test/test_helper.rb
ADDED
metadata
ADDED
@@ -0,0 +1,171 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: fluent-plugin-redshift-out2
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: 0.4.1
|
5
|
+
platform: ruby
|
6
|
+
authors:
|
7
|
+
- Ertugrul Yilmaz
|
8
|
+
autorequire:
|
9
|
+
bindir: bin
|
10
|
+
cert_chain: []
|
11
|
+
date: 2020-07-09 00:00:00.000000000 Z
|
12
|
+
dependencies:
|
13
|
+
- !ruby/object:Gem::Dependency
|
14
|
+
name: fluentd
|
15
|
+
requirement: !ruby/object:Gem::Requirement
|
16
|
+
requirements:
|
17
|
+
- - ">="
|
18
|
+
- !ruby/object:Gem::Version
|
19
|
+
version: 0.10.0
|
20
|
+
- - "<"
|
21
|
+
- !ruby/object:Gem::Version
|
22
|
+
version: '2'
|
23
|
+
type: :runtime
|
24
|
+
prerelease: false
|
25
|
+
version_requirements: !ruby/object:Gem::Requirement
|
26
|
+
requirements:
|
27
|
+
- - ">="
|
28
|
+
- !ruby/object:Gem::Version
|
29
|
+
version: 0.10.0
|
30
|
+
- - "<"
|
31
|
+
- !ruby/object:Gem::Version
|
32
|
+
version: '2'
|
33
|
+
- !ruby/object:Gem::Dependency
|
34
|
+
name: aws-sdk-v1
|
35
|
+
requirement: !ruby/object:Gem::Requirement
|
36
|
+
requirements:
|
37
|
+
- - ">="
|
38
|
+
- !ruby/object:Gem::Version
|
39
|
+
version: 1.6.3
|
40
|
+
type: :runtime
|
41
|
+
prerelease: false
|
42
|
+
version_requirements: !ruby/object:Gem::Requirement
|
43
|
+
requirements:
|
44
|
+
- - ">="
|
45
|
+
- !ruby/object:Gem::Version
|
46
|
+
version: 1.6.3
|
47
|
+
- !ruby/object:Gem::Dependency
|
48
|
+
name: multi_json
|
49
|
+
requirement: !ruby/object:Gem::Requirement
|
50
|
+
requirements:
|
51
|
+
- - "~>"
|
52
|
+
- !ruby/object:Gem::Version
|
53
|
+
version: '1.10'
|
54
|
+
type: :runtime
|
55
|
+
prerelease: false
|
56
|
+
version_requirements: !ruby/object:Gem::Requirement
|
57
|
+
requirements:
|
58
|
+
- - "~>"
|
59
|
+
- !ruby/object:Gem::Version
|
60
|
+
version: '1.10'
|
61
|
+
- !ruby/object:Gem::Dependency
|
62
|
+
name: yajl-ruby
|
63
|
+
requirement: !ruby/object:Gem::Requirement
|
64
|
+
requirements:
|
65
|
+
- - "~>"
|
66
|
+
- !ruby/object:Gem::Version
|
67
|
+
version: '1.2'
|
68
|
+
type: :runtime
|
69
|
+
prerelease: false
|
70
|
+
version_requirements: !ruby/object:Gem::Requirement
|
71
|
+
requirements:
|
72
|
+
- - "~>"
|
73
|
+
- !ruby/object:Gem::Version
|
74
|
+
version: '1.2'
|
75
|
+
- !ruby/object:Gem::Dependency
|
76
|
+
name: pg
|
77
|
+
requirement: !ruby/object:Gem::Requirement
|
78
|
+
requirements:
|
79
|
+
- - "~>"
|
80
|
+
- !ruby/object:Gem::Version
|
81
|
+
version: 0.17.0
|
82
|
+
type: :runtime
|
83
|
+
prerelease: false
|
84
|
+
version_requirements: !ruby/object:Gem::Requirement
|
85
|
+
requirements:
|
86
|
+
- - "~>"
|
87
|
+
- !ruby/object:Gem::Version
|
88
|
+
version: 0.17.0
|
89
|
+
- !ruby/object:Gem::Dependency
|
90
|
+
name: rake
|
91
|
+
requirement: !ruby/object:Gem::Requirement
|
92
|
+
requirements:
|
93
|
+
- - ">="
|
94
|
+
- !ruby/object:Gem::Version
|
95
|
+
version: '0'
|
96
|
+
type: :development
|
97
|
+
prerelease: false
|
98
|
+
version_requirements: !ruby/object:Gem::Requirement
|
99
|
+
requirements:
|
100
|
+
- - ">="
|
101
|
+
- !ruby/object:Gem::Version
|
102
|
+
version: '0'
|
103
|
+
- !ruby/object:Gem::Dependency
|
104
|
+
name: simplecov
|
105
|
+
requirement: !ruby/object:Gem::Requirement
|
106
|
+
requirements:
|
107
|
+
- - ">="
|
108
|
+
- !ruby/object:Gem::Version
|
109
|
+
version: 0.5.4
|
110
|
+
type: :development
|
111
|
+
prerelease: false
|
112
|
+
version_requirements: !ruby/object:Gem::Requirement
|
113
|
+
requirements:
|
114
|
+
- - ">="
|
115
|
+
- !ruby/object:Gem::Version
|
116
|
+
version: 0.5.4
|
117
|
+
- !ruby/object:Gem::Dependency
|
118
|
+
name: flexmock
|
119
|
+
requirement: !ruby/object:Gem::Requirement
|
120
|
+
requirements:
|
121
|
+
- - ">="
|
122
|
+
- !ruby/object:Gem::Version
|
123
|
+
version: 1.3.1
|
124
|
+
type: :development
|
125
|
+
prerelease: false
|
126
|
+
version_requirements: !ruby/object:Gem::Requirement
|
127
|
+
requirements:
|
128
|
+
- - ">="
|
129
|
+
- !ruby/object:Gem::Version
|
130
|
+
version: 1.3.1
|
131
|
+
description: Amazon Redshift output plugin for Fluentd
|
132
|
+
email:
|
133
|
+
- "***.***.***@gmail.com"
|
134
|
+
executables: []
|
135
|
+
extensions: []
|
136
|
+
extra_rdoc_files: []
|
137
|
+
files:
|
138
|
+
- ".gitignore"
|
139
|
+
- Gemfile
|
140
|
+
- README.md
|
141
|
+
- Rakefile
|
142
|
+
- VERSION
|
143
|
+
- fluent-plugin-redshift-out.gemspec
|
144
|
+
- lib/fluent/plugin/out_redshift-out.rb
|
145
|
+
- test/plugin/test_out_redshift.rb
|
146
|
+
- test/test_helper.rb
|
147
|
+
homepage: https://github.com/ertugrulyilmaz/fluent-plugin-redshift-out
|
148
|
+
licenses: []
|
149
|
+
metadata: {}
|
150
|
+
post_install_message:
|
151
|
+
rdoc_options: []
|
152
|
+
require_paths:
|
153
|
+
- lib
|
154
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
155
|
+
requirements:
|
156
|
+
- - ">="
|
157
|
+
- !ruby/object:Gem::Version
|
158
|
+
version: '0'
|
159
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
160
|
+
requirements:
|
161
|
+
- - ">="
|
162
|
+
- !ruby/object:Gem::Version
|
163
|
+
version: '0'
|
164
|
+
requirements: []
|
165
|
+
rubygems_version: 3.0.3
|
166
|
+
signing_key:
|
167
|
+
specification_version: 4
|
168
|
+
summary: Amazon Redshift output plugin for Fluentd
|
169
|
+
test_files:
|
170
|
+
- test/plugin/test_out_redshift.rb
|
171
|
+
- test/test_helper.rb
|