taps 0.3.15 → 0.3.17
Sign up to get free protection for your applications and to get access to all the features.
- data/VERSION.yml +1 -1
- data/bin/schema +1 -1
- data/lib/taps/chunksize.rb +52 -0
- data/lib/taps/cli.rb +2 -0
- data/lib/taps/data_stream.rb +11 -1
- data/lib/taps/operation.rb +43 -23
- data/lib/taps/utils.rb +11 -28
- data/spec/chunksize_spec.rb +41 -0
- data/spec/cli_spec.rb +6 -0
- data/spec/operation_spec.rb +10 -0
- data/spec/utils_spec.rb +0 -31
- metadata +13 -11
data/VERSION.yml
CHANGED
data/bin/schema
CHANGED
@@ -0,0 +1,52 @@
|
|
1
|
+
require 'taps/errors'
|
2
|
+
|
3
|
+
class Taps::Chunksize
|
4
|
+
attr_accessor :idle_secs, :time_in_db, :start_time, :end_time, :retries
|
5
|
+
attr_reader :chunksize
|
6
|
+
|
7
|
+
def initialize(chunksize)
|
8
|
+
@chunksize = chunksize
|
9
|
+
@idle_secs = 0.0
|
10
|
+
@retries = 0
|
11
|
+
end
|
12
|
+
|
13
|
+
def to_i
|
14
|
+
chunksize
|
15
|
+
end
|
16
|
+
|
17
|
+
def reset_chunksize
|
18
|
+
@chunksize = (retries <= 1) ? 10 : 1
|
19
|
+
end
|
20
|
+
|
21
|
+
def diff
|
22
|
+
end_time - start_time - time_in_db - idle_secs
|
23
|
+
end
|
24
|
+
|
25
|
+
def time_in_db=(t)
|
26
|
+
@time_in_db = t
|
27
|
+
@time_in_db = @time_in_db.to_f rescue 0.0
|
28
|
+
end
|
29
|
+
|
30
|
+
def time_delta
|
31
|
+
t1 = Time.now
|
32
|
+
yield if block_given?
|
33
|
+
t2 = Time.now
|
34
|
+
t2 - t1
|
35
|
+
end
|
36
|
+
|
37
|
+
def calc_new_chunksize
|
38
|
+
new_chunksize = if retries > 0
|
39
|
+
chunksize
|
40
|
+
elsif diff > 3.0
|
41
|
+
(chunksize / 3).ceil
|
42
|
+
elsif diff > 1.1
|
43
|
+
chunksize - 100
|
44
|
+
elsif diff < 0.8
|
45
|
+
chunksize * 2
|
46
|
+
else
|
47
|
+
chunksize + 100
|
48
|
+
end
|
49
|
+
new_chunksize = 1 if new_chunksize < 1
|
50
|
+
new_chunksize
|
51
|
+
end
|
52
|
+
end
|
data/lib/taps/cli.rb
CHANGED
@@ -128,6 +128,7 @@ EOHELP
|
|
128
128
|
o.define_head "Push a database to a taps server"
|
129
129
|
end
|
130
130
|
|
131
|
+
o.on("-s", "--skip-schema", "Don't transfer the schema, just data") { |v| opts[:skip_schema] = true }
|
131
132
|
o.on("-i", "--indexes-first", "Transfer indexes first before data") { |v| opts[:indexes_first] = true }
|
132
133
|
o.on("-r", "--resume=file", "Resume a Taps Session from a stored file") { |v| opts[:resume_filename] = v }
|
133
134
|
o.on("-c", "--chunksize=N", "Initial Chunksize") { |v| opts[:default_chunksize] = (v.to_i < 10 ? 10 : v.to_i) }
|
@@ -137,6 +138,7 @@ EOHELP
|
|
137
138
|
r_tables = v.collect { |t| "^#{t}$" }.join("|")
|
138
139
|
opts[:table_filter] = "(#{r_tables})"
|
139
140
|
end
|
141
|
+
o.on("-e", "--exclude_tables=A,B,C", Array, "Shortcut to exclude a list of tables") { |v| opts[:exclude_tables] = v }
|
140
142
|
o.on("-d", "--debug", "Enable Debug Messages") { |v| opts[:debug] = true }
|
141
143
|
o.parse!(argv)
|
142
144
|
|
data/lib/taps/data_stream.rb
CHANGED
@@ -166,7 +166,7 @@ class DataStream
|
|
166
166
|
res = nil
|
167
167
|
log.debug "DataStream#fetch_from_resource state -> #{state.inspect}"
|
168
168
|
state[:chunksize] = Taps::Utils.calculate_chunksize(state[:chunksize]) do |c|
|
169
|
-
state[:chunksize] = c
|
169
|
+
state[:chunksize] = c.to_i
|
170
170
|
res = resource.post({:state => self.to_json}, headers)
|
171
171
|
end
|
172
172
|
|
@@ -202,6 +202,16 @@ class DataStream
|
|
202
202
|
def import_rows(rows)
|
203
203
|
table.import(rows[:header], rows[:data])
|
204
204
|
state[:offset] += rows[:data].size
|
205
|
+
rescue Exception => ex
|
206
|
+
case ex.message
|
207
|
+
when /integer out of range/ then
|
208
|
+
raise Taps::InvalidData, <<-ERROR, []
|
209
|
+
\nDetected integer data that exceeds the maximum allowable size for an integer type.
|
210
|
+
This generally occurs when importing from SQLite due to the fact that SQLite does
|
211
|
+
not enforce maximum values on integer types.
|
212
|
+
ERROR
|
213
|
+
else raise ex
|
214
|
+
end
|
205
215
|
end
|
206
216
|
|
207
217
|
def verify_stream
|
data/lib/taps/operation.rb
CHANGED
@@ -29,6 +29,10 @@ class Operation
|
|
29
29
|
"op"
|
30
30
|
end
|
31
31
|
|
32
|
+
def skip_schema?
|
33
|
+
!!opts[:skip_schema]
|
34
|
+
end
|
35
|
+
|
32
36
|
def indexes_first?
|
33
37
|
!!opts[:indexes_first]
|
34
38
|
end
|
@@ -37,19 +41,24 @@ class Operation
|
|
37
41
|
opts[:table_filter]
|
38
42
|
end
|
39
43
|
|
44
|
+
def exclude_tables
|
45
|
+
opts[:exclude_tables] || []
|
46
|
+
end
|
47
|
+
|
40
48
|
def apply_table_filter(tables)
|
41
|
-
return tables unless table_filter
|
42
|
-
|
49
|
+
return tables unless table_filter || exclude_tables
|
50
|
+
|
51
|
+
re = table_filter ? Regexp.new(table_filter) : nil
|
43
52
|
if tables.kind_of?(Hash)
|
44
53
|
ntables = {}
|
45
54
|
tables.each do |t, d|
|
46
|
-
|
55
|
+
if !exclude_tables.include?(t.to_s) && (!re || !re.match(t.to_s).nil?)
|
47
56
|
ntables[t] = d
|
48
57
|
end
|
49
58
|
end
|
50
59
|
ntables
|
51
60
|
else
|
52
|
-
tables.reject { |t| re.match(t.to_s).nil? }
|
61
|
+
tables.reject { |t| exclude_tables.include?(t.to_s) || (re && re.match(t.to_s).nil?) }
|
53
62
|
end
|
54
63
|
end
|
55
64
|
|
@@ -236,13 +245,13 @@ class Pull < Operation
|
|
236
245
|
def run
|
237
246
|
catch_errors do
|
238
247
|
unless resuming?
|
239
|
-
pull_schema
|
240
|
-
pull_indexes if indexes_first?
|
248
|
+
pull_schema if !skip_schema?
|
249
|
+
pull_indexes if indexes_first? && !skip_schema?
|
241
250
|
end
|
242
251
|
setup_signal_trap
|
243
252
|
pull_partial_data if resuming?
|
244
253
|
pull_data
|
245
|
-
pull_indexes
|
254
|
+
pull_indexes if !indexes_first? && !skip_schema?
|
246
255
|
pull_reset_sequences
|
247
256
|
end
|
248
257
|
end
|
@@ -395,13 +404,13 @@ class Push < Operation
|
|
395
404
|
def run
|
396
405
|
catch_errors do
|
397
406
|
unless resuming?
|
398
|
-
push_schema
|
399
|
-
push_indexes if indexes_first?
|
407
|
+
push_schema if !skip_schema?
|
408
|
+
push_indexes if indexes_first? && !skip_schema?
|
400
409
|
end
|
401
410
|
setup_signal_trap
|
402
411
|
push_partial_data if resuming?
|
403
412
|
push_data
|
404
|
-
push_indexes
|
413
|
+
push_indexes if !indexes_first? && !skip_schema?
|
405
414
|
push_reset_sequences
|
406
415
|
end
|
407
416
|
end
|
@@ -480,23 +489,32 @@ class Push < Operation
|
|
480
489
|
|
481
490
|
begin
|
482
491
|
chunksize = Taps::Utils.calculate_chunksize(chunksize) do |c|
|
483
|
-
stream.state[:chunksize] = c
|
484
|
-
encoded_data, row_size, elapsed_time =
|
492
|
+
stream.state[:chunksize] = c.to_i
|
493
|
+
encoded_data, row_size, elapsed_time = nil
|
494
|
+
d1 = c.time_delta do
|
495
|
+
encoded_data, row_size, elapsed_time = stream.fetch
|
496
|
+
end
|
485
497
|
break if stream.complete?
|
486
498
|
|
487
|
-
data =
|
488
|
-
|
489
|
-
|
490
|
-
|
499
|
+
data = nil
|
500
|
+
d2 = c.time_delta do
|
501
|
+
data = {
|
502
|
+
:state => stream.to_hash,
|
503
|
+
:checksum => Taps::Utils.checksum(encoded_data).to_s
|
504
|
+
}
|
505
|
+
end
|
491
506
|
|
492
507
|
begin
|
493
|
-
content, content_type =
|
494
|
-
|
495
|
-
|
496
|
-
:
|
497
|
-
|
498
|
-
|
499
|
-
:
|
508
|
+
content, content_type = nil
|
509
|
+
d3 = c.time_delta do
|
510
|
+
content, content_type = Taps::Multipart.create do |r|
|
511
|
+
r.attach :name => :encoded_data,
|
512
|
+
:payload => encoded_data,
|
513
|
+
:content_type => 'application/octet-stream'
|
514
|
+
r.attach :name => :json,
|
515
|
+
:payload => data.to_json,
|
516
|
+
:content_type => 'application/json'
|
517
|
+
end
|
500
518
|
end
|
501
519
|
session_resource['push/table'].post(content, http_headers(:content_type => content_type))
|
502
520
|
self.stream_state = stream.to_hash
|
@@ -504,6 +522,8 @@ class Push < Operation
|
|
504
522
|
Taps::Utils.reraise_server_exception(e)
|
505
523
|
end
|
506
524
|
|
525
|
+
c.idle_secs = (d1 + d2 + d3)
|
526
|
+
|
507
527
|
elapsed_time
|
508
528
|
end
|
509
529
|
rescue Taps::CorruptedData => e
|
data/lib/taps/utils.rb
CHANGED
@@ -2,8 +2,10 @@ require 'zlib'
|
|
2
2
|
require 'stringio'
|
3
3
|
require 'time'
|
4
4
|
require 'tempfile'
|
5
|
+
require 'rest_client'
|
5
6
|
|
6
7
|
require 'taps/errors'
|
8
|
+
require 'taps/chunksize'
|
7
9
|
|
8
10
|
module Taps
|
9
11
|
module Utils
|
@@ -93,42 +95,23 @@ Data : #{data}
|
|
93
95
|
end
|
94
96
|
|
95
97
|
def calculate_chunksize(old_chunksize)
|
96
|
-
|
98
|
+
c = Taps::Chunksize.new(old_chunksize)
|
97
99
|
|
98
|
-
retries = 0
|
99
|
-
time_in_db = 0
|
100
100
|
begin
|
101
|
-
|
102
|
-
time_in_db = yield
|
103
|
-
time_in_db = time_in_db.to_f rescue 0
|
101
|
+
c.start_time = Time.now
|
102
|
+
c.time_in_db = yield c
|
104
103
|
rescue Errno::EPIPE, RestClient::RequestFailed, RestClient::RequestTimeout
|
105
|
-
retries += 1
|
106
|
-
raise if retries > 2
|
104
|
+
c.retries += 1
|
105
|
+
raise if c.retries > 2
|
107
106
|
|
108
107
|
# we got disconnected, the chunksize could be too large
|
109
|
-
#
|
110
|
-
|
111
|
-
|
108
|
+
# reset the chunksize based on the number of retries
|
109
|
+
c.reset_chunksize
|
112
110
|
retry
|
113
111
|
end
|
114
112
|
|
115
|
-
|
116
|
-
|
117
|
-
diff = t2 - t1 - time_in_db
|
118
|
-
|
119
|
-
new_chunksize = if retries > 0
|
120
|
-
chunksize
|
121
|
-
elsif diff > 3.0
|
122
|
-
(chunksize / 3).ceil
|
123
|
-
elsif diff > 1.1
|
124
|
-
chunksize - 100
|
125
|
-
elsif diff < 0.8
|
126
|
-
chunksize * 2
|
127
|
-
else
|
128
|
-
chunksize + 100
|
129
|
-
end
|
130
|
-
new_chunksize = 1 if new_chunksize < 1
|
131
|
-
new_chunksize
|
113
|
+
c.end_time = Time.now
|
114
|
+
c.calc_new_chunksize
|
132
115
|
end
|
133
116
|
|
134
117
|
def load_schema(database_url, schema_data)
|
@@ -0,0 +1,41 @@
|
|
1
|
+
require File.dirname(__FILE__) + '/base'
|
2
|
+
require 'taps/utils'
|
3
|
+
|
4
|
+
describe Taps::Chunksize do
|
5
|
+
it "scales chunksize down slowly when the time delta of the block is just over a second" do
|
6
|
+
Time.stubs(:now).returns(10.0).returns(11.5)
|
7
|
+
Taps::Utils.calculate_chunksize(1000) { |c| }.should == 900
|
8
|
+
end
|
9
|
+
|
10
|
+
it "scales chunksize down fast when the time delta of the block is over 3 seconds" do
|
11
|
+
Time.stubs(:now).returns(10.0).returns(15.0)
|
12
|
+
Taps::Utils.calculate_chunksize(3000) { |c| }.should == 1000
|
13
|
+
end
|
14
|
+
|
15
|
+
it "scales up chunksize fast when the time delta of the block is under 0.8 seconds" do
|
16
|
+
Time.stubs(:now).returns(10.0).returns(10.7)
|
17
|
+
Taps::Utils.calculate_chunksize(1000) { |c| }.should == 2000
|
18
|
+
end
|
19
|
+
|
20
|
+
it "scales up chunksize slow when the time delta of the block is between 0.8 and 1.1 seconds" do
|
21
|
+
Time.stubs(:now).returns(10.0).returns(10.8)
|
22
|
+
Taps::Utils.calculate_chunksize(1000) { |c| }.should == 1100
|
23
|
+
|
24
|
+
Time.stubs(:now).returns(10.0).returns(11.1)
|
25
|
+
Taps::Utils.calculate_chunksize(1000) { |c| }.should == 1100
|
26
|
+
end
|
27
|
+
|
28
|
+
it "will reset the chunksize to a small value if we got a broken pipe exception" do
|
29
|
+
Taps::Utils.calculate_chunksize(1000) do |c|
|
30
|
+
raise Errno::EPIPE if c.chunksize == 1000
|
31
|
+
c.chunksize.should == 10
|
32
|
+
end.should == 10
|
33
|
+
end
|
34
|
+
|
35
|
+
it "will reset the chunksize to a small value if we got a broken pipe exception a second time" do
|
36
|
+
Taps::Utils.calculate_chunksize(1000) do |c|
|
37
|
+
raise Errno::EPIPE if c.chunksize == 1000 || c.chunksize == 10
|
38
|
+
c.chunksize.should == 1
|
39
|
+
end.should == 1
|
40
|
+
end
|
41
|
+
end
|
data/spec/cli_spec.rb
CHANGED
@@ -7,4 +7,10 @@ describe Taps::Cli do
|
|
7
7
|
opts = @cli.clientoptparse(:pull)
|
8
8
|
opts[:table_filter].should == "(^mytable1$|^logs$)"
|
9
9
|
end
|
10
|
+
|
11
|
+
it "translates a list of tables to exclude into a regex that can be used in table_filter" do
|
12
|
+
@cli = Taps::Cli.new(["-e", "mytable1,logs", "sqlite://tmp.db", "http://x:y@localhost:5000"])
|
13
|
+
opts = @cli.clientoptparse(:pull)
|
14
|
+
opts[:exclude_tables].should == ['mytable1','logs']
|
15
|
+
end
|
10
16
|
end
|
data/spec/operation_spec.rb
CHANGED
@@ -16,6 +16,16 @@ describe Taps::Operation do
|
|
16
16
|
@op.apply_table_filter({ 'abc' => 1, 'def' => 2 }).should == { 'abc' => 1 }
|
17
17
|
end
|
18
18
|
|
19
|
+
it "returns an array of tables without the exclude_tables tables" do
|
20
|
+
@op = Taps::Operation.new('dummy://localhost', 'http://x:y@localhost:5000', :exclude_tables => ['abc', 'ghi', 'jkl'])
|
21
|
+
@op.apply_table_filter(['abc', 'def', 'ghi', 'jkl', 'mno']).should == ['def', 'mno']
|
22
|
+
end
|
23
|
+
|
24
|
+
it "returns a hash of tables without the exclude_tables tables" do
|
25
|
+
@op = Taps::Operation.new('dummy://localhost', 'http://x:y@localhost:5000', :exclude_tables => ['abc', 'ghi', 'jkl'])
|
26
|
+
@op.apply_table_filter({ 'abc' => 1, 'def' => 2, 'ghi' => 3, 'jkl' => 4, 'mno' => 5 }).should == { 'def' => 2, 'mno' => 5 }
|
27
|
+
end
|
28
|
+
|
19
29
|
it "masks a url's password" do
|
20
30
|
@op.safe_url("mysql://root:password@localhost/mydb").should == "mysql://root:[hidden]@localhost/mydb"
|
21
31
|
end
|
data/spec/utils_spec.rb
CHANGED
@@ -18,37 +18,6 @@ describe Taps::Utils do
|
|
18
18
|
lambda { Taps::Utils.format_data(data, :schema => schema) }.should.raise(Taps::InvalidData)
|
19
19
|
end
|
20
20
|
|
21
|
-
it "scales chunksize down slowly when the time delta of the block is just over a second" do
|
22
|
-
Time.stubs(:now).returns(10.0).returns(11.5)
|
23
|
-
Taps::Utils.calculate_chunksize(1000) { }.should == 900
|
24
|
-
end
|
25
|
-
|
26
|
-
it "scales chunksize down fast when the time delta of the block is over 3 seconds" do
|
27
|
-
Time.stubs(:now).returns(10.0).returns(15.0)
|
28
|
-
Taps::Utils.calculate_chunksize(3000) { }.should == 1000
|
29
|
-
end
|
30
|
-
|
31
|
-
it "scales up chunksize fast when the time delta of the block is under 0.8 seconds" do
|
32
|
-
Time.stubs(:now).returns(10.0).returns(10.7)
|
33
|
-
Taps::Utils.calculate_chunksize(1000) { }.should == 2000
|
34
|
-
end
|
35
|
-
|
36
|
-
it "scales up chunksize slow when the time delta of the block is between 0.8 and 1.1 seconds" do
|
37
|
-
Time.stubs(:now).returns(10.0).returns(10.8)
|
38
|
-
Taps::Utils.calculate_chunksize(1000) { }.should == 1100
|
39
|
-
|
40
|
-
Time.stubs(:now).returns(10.0).returns(11.1)
|
41
|
-
Taps::Utils.calculate_chunksize(1000) { }.should == 1100
|
42
|
-
end
|
43
|
-
|
44
|
-
it "will reset the chunksize to a small value if we got a broken pipe exception" do
|
45
|
-
Taps::Utils.calculate_chunksize(1000) { |c| raise Errno::EPIPE if c == 1000; c.should == 10 }.should == 10
|
46
|
-
end
|
47
|
-
|
48
|
-
it "will reset the chunksize to a small value if we got a broken pipe exception a second time" do
|
49
|
-
Taps::Utils.calculate_chunksize(1000) { |c| raise Errno::EPIPE if c == 1000 || c == 10; c.should == 1 }.should == 1
|
50
|
-
end
|
51
|
-
|
52
21
|
it "returns a list of columns that are text fields if the database is mysql" do
|
53
22
|
@db = mock("db", :url => "mysql://localhost/mydb")
|
54
23
|
@db.stubs(:schema).with(:mytable).returns([
|
metadata
CHANGED
@@ -1,13 +1,13 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: taps
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
hash:
|
4
|
+
hash: 49
|
5
5
|
prerelease:
|
6
6
|
segments:
|
7
7
|
- 0
|
8
8
|
- 3
|
9
|
-
-
|
10
|
-
version: 0.3.
|
9
|
+
- 17
|
10
|
+
version: 0.3.17
|
11
11
|
platform: ruby
|
12
12
|
authors:
|
13
13
|
- Ricardo Chimal, Jr.
|
@@ -15,7 +15,7 @@ autorequire:
|
|
15
15
|
bindir: bin
|
16
16
|
cert_chain: []
|
17
17
|
|
18
|
-
date: 2011-
|
18
|
+
date: 2011-02-25 00:00:00 -08:00
|
19
19
|
default_executable:
|
20
20
|
dependencies:
|
21
21
|
- !ruby/object:Gem::Dependency
|
@@ -26,12 +26,12 @@ dependencies:
|
|
26
26
|
requirements:
|
27
27
|
- - ~>
|
28
28
|
- !ruby/object:Gem::Version
|
29
|
-
hash:
|
29
|
+
hash: 1
|
30
30
|
segments:
|
31
31
|
- 1
|
32
|
-
-
|
33
|
-
-
|
34
|
-
version: 1.
|
32
|
+
- 5
|
33
|
+
- 1
|
34
|
+
version: 1.5.1
|
35
35
|
type: :runtime
|
36
36
|
version_requirements: *id001
|
37
37
|
- !ruby/object:Gem::Dependency
|
@@ -82,12 +82,12 @@ dependencies:
|
|
82
82
|
requirements:
|
83
83
|
- - ~>
|
84
84
|
- !ruby/object:Gem::Version
|
85
|
-
hash:
|
85
|
+
hash: 87
|
86
86
|
segments:
|
87
87
|
- 3
|
88
|
-
-
|
88
|
+
- 20
|
89
89
|
- 0
|
90
|
-
version: 3.
|
90
|
+
version: 3.20.0
|
91
91
|
type: :runtime
|
92
92
|
version_requirements: *id004
|
93
93
|
- !ruby/object:Gem::Dependency
|
@@ -204,6 +204,7 @@ files:
|
|
204
204
|
- bin/schema
|
205
205
|
- bin/schema.cmd
|
206
206
|
- bin/taps
|
207
|
+
- lib/taps/chunksize.rb
|
207
208
|
- lib/taps/cli.rb
|
208
209
|
- lib/taps/config.rb
|
209
210
|
- lib/taps/data_stream.rb
|
@@ -220,6 +221,7 @@ files:
|
|
220
221
|
- lib/taps/version.rb
|
221
222
|
- README.rdoc
|
222
223
|
- spec/base.rb
|
224
|
+
- spec/chunksize_spec.rb
|
223
225
|
- spec/cli_spec.rb
|
224
226
|
- spec/data_stream_spec.rb
|
225
227
|
- spec/operation_spec.rb
|