Class: WebHDFS::ClientV1

Inherits:
Object
  • Object
show all
Defined in:
lib/webhdfs/client_v1.rb

Direct Known Subclasses

Client

Constant Summary collapse

OPT_TABLE =

This hash table holds command options.

{}
KNOWN_ERRORS =

internal use only

['LeaseExpiredException'].freeze
SSL_VERIFY_MODES = 
[:none, :peer]
REDIRECTED_OPERATIONS = 
['APPEND', 'CREATE', 'OPEN', 'GETFILECHECKSUM']

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(host = 'localhost', port = 50070, username = nil, doas = nil, proxy_address = nil, proxy_port = nil) ⇒ ClientV1

Returns a new instance of ClientV1.



36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
 
# File 'lib/webhdfs/client_v1.rb', line 36

def initialize(host='localhost', port=50070, username=nil, doas=nil, proxy_address=nil, proxy_port=nil)
  @host = host
  @port = port
  @username = username
  @doas = doas
  @proxy_address = proxy_address
  @proxy_port = proxy_port
  @retry_known_errors = false
  @retry_times = 1
  @retry_interval = 1

  @httpfs_mode = false

  @ssl = false
  @ssl_ca_file = nil
  @ssl_verify_mode = nil

  @kerberos = false
end

Instance Attribute Details

#doas ⇒ Object

Returns the value of attribute doas.



15
16
17
 
# File 'lib/webhdfs/client_v1.rb', line 15

def doas
  @doas
end

#host ⇒ Object

Returns the value of attribute host.



15
16
17
 
# File 'lib/webhdfs/client_v1.rb', line 15

def host
  @host
end

#httpfs_mode ⇒ Object

Returns the value of attribute httpfs_mode.



19
20
21
 
# File 'lib/webhdfs/client_v1.rb', line 19

def httpfs_mode
  @httpfs_mode
end

#kerberos ⇒ Object

Returns the value of attribute kerberos.



26
27
28
 
# File 'lib/webhdfs/client_v1.rb', line 26

def kerberos
  @kerberos
end

#open_timeout ⇒ Object

default 30s (in ruby net/http)



17
18
19
 
# File 'lib/webhdfs/client_v1.rb', line 17

def open_timeout
  @open_timeout
end

#port ⇒ Object

Returns the value of attribute port.



15
16
17
 
# File 'lib/webhdfs/client_v1.rb', line 15

def port
  @port
end

#proxy_address ⇒ Object

Returns the value of attribute proxy_address.



15
16
17
 
# File 'lib/webhdfs/client_v1.rb', line 15

def proxy_address
  @proxy_address
end

#proxy_pass ⇒ Object

Returns the value of attribute proxy_pass.



16
17
18
 
# File 'lib/webhdfs/client_v1.rb', line 16

def proxy_pass
  @proxy_pass
end

#proxy_port ⇒ Object

Returns the value of attribute proxy_port.



15
16
17
 
# File 'lib/webhdfs/client_v1.rb', line 15

def proxy_port
  @proxy_port
end

#proxy_user ⇒ Object

Returns the value of attribute proxy_user.



16
17
18
 
# File 'lib/webhdfs/client_v1.rb', line 16

def proxy_user
  @proxy_user
end

#read_timeout ⇒ Object

default 60s (in ruby net/http)



18
19
20
 
# File 'lib/webhdfs/client_v1.rb', line 18

def read_timeout
  @read_timeout
end

#retry_interval ⇒ Object

default 1 ([sec], ignored when retry_known_errors is false)



22
23
24
 
# File 'lib/webhdfs/client_v1.rb', line 22

def retry_interval
  @retry_interval
end

#retry_known_errors ⇒ Object

default false (not to retry)



20
21
22
 
# File 'lib/webhdfs/client_v1.rb', line 20

def retry_known_errors
  @retry_known_errors
end

#retry_times ⇒ Object

default 1 (ignored when retry_known_errors is false)



21
22
23
 
# File 'lib/webhdfs/client_v1.rb', line 21

def retry_times
  @retry_times
end

#ssl ⇒ Object

Returns the value of attribute ssl.



23
24
25
 
# File 'lib/webhdfs/client_v1.rb', line 23

def ssl
  @ssl
end

#ssl_ca_file ⇒ Object

Returns the value of attribute ssl_ca_file.



24
25
26
 
# File 'lib/webhdfs/client_v1.rb', line 24

def ssl_ca_file
  @ssl_ca_file
end

#ssl_verify_mode ⇒ Object

Returns the value of attribute ssl_verify_mode.



25
26
27
 
# File 'lib/webhdfs/client_v1.rb', line 25

def ssl_verify_mode
  @ssl_verify_mode
end

#username ⇒ Object

Returns the value of attribute username.



15
16
17
 
# File 'lib/webhdfs/client_v1.rb', line 15

def username
  @username
end

Instance Method Details

#api_path(path) ⇒ Object 



225
226
227
228
229
230
231
 
# File 'lib/webhdfs/client_v1.rb', line 225

def api_path(path)
  if path.start_with?('/')
    '/webhdfs/v1' + path
  else
    '/webhdfs/v1/' + path
  end
end

#append(path, body, options = {}) ⇒ Object

curl -i -X POST “<HOST>:<PORT>/webhdfs/v1/<PATH>?op=APPEND”



70
71
72
73
74
75
76
77
 
# File 'lib/webhdfs/client_v1.rb', line 70

def append(path, body, options={})
  if @httpfs_mode
    options = options.merge({'data' => 'true'})
  end
  check_options(options, OPT_TABLE['APPEND'])
  res = operate_requests('POST', path, 'APPEND', options, body)
  res.code == '200'
end

#build_path(path, op, params) ⇒ Object 



233
234
235
236
237
238
239
240
241
242
243
244
245
 
# File 'lib/webhdfs/client_v1.rb', line 233

def build_path(path, op, params)
  opts = if @username and @doas
           {'op' => op, 'user.name' => @username, 'doas' => @doas}
         elsif @username
           {'op' => op, 'user.name' => @username}
         elsif @doas
           {'op' => op, 'doas' => @doas}
         else
           {'op' => op}
         end
  query = URI.encode_www_form(params.merge(opts))
  api_path(path) + '?' + query
end

#check_options(options, optdecl = []) ⇒ Object

def delegation_token(user, options={}) # GETDELEGATIONTOKEN

raise NotImplementedError

end def renew_delegation_token(token, options={}) # RENEWDELEGATIONTOKEN

raise NotImplementedError

end def cancel_delegation_token(token, options={}) # CANCELDELEGATIONTOKEN

raise NotImplementedError

end

Raises:

  • (ArgumentError) 


216
217
218
219
 
# File 'lib/webhdfs/client_v1.rb', line 216

def check_options(options, optdecl=[])
  ex = options.keys.map(&:to_s) - (optdecl || [])
  raise ArgumentError, "no such option: #{ex.join(' ')}" unless ex.empty?
end

#check_success_json(res, attr = nil) ⇒ Object 



221
222
223
 
# File 'lib/webhdfs/client_v1.rb', line 221

def check_success_json(res, attr=nil)
  res.code == '200' and res.content_type == 'application/json' and (attr.nil? or JSON.parse(res.body)[attr])
end

#checksum(path, options = {}) ⇒ Object Also known as: getfilechecksum

curl -i “http://<HOST>:<PORT>/webhdfs/v1/<PATH>?op=GETFILECHECKSUM”



143
144
145
146
147
 
# File 'lib/webhdfs/client_v1.rb', line 143

def checksum(path, options={})
  check_options(options, OPT_TABLE['GETFILECHECKSUM'])
  res = operate_requests('GET', path, 'GETFILECHECKSUM', options)
  check_success_json(res, 'FileChecksum')
end

#chmod(path, mode, options = {}) ⇒ Object Also known as: setpermission

curl -i -X PUT “http://<HOST>:<PORT>/webhdfs/v1/<PATH>?op=SETPERMISSION

[&permission=<OCTAL>]"



160
161
162
163
164
 
# File 'lib/webhdfs/client_v1.rb', line 160

def chmod(path, mode, options={})
  check_options(options, OPT_TABLE['SETPERMISSION'])
  res = operate_requests('PUT', path, 'SETPERMISSION', options.merge({'permission' => mode}))
  res.code == '200'
end

#chown(path, options = {}) ⇒ Object Also known as: setowner

curl -i -X PUT “http://<HOST>:<PORT>/webhdfs/v1/<PATH>?op=SETOWNER

[&owner=<USER>][&group=<GROUP>]"



169
170
171
172
173
174
175
176
177
 
# File 'lib/webhdfs/client_v1.rb', line 169

def chown(path, options={})
  check_options(options, OPT_TABLE['SETOWNER'])
  unless options.has_key?('owner') or options.has_key?('group') or
      options.has_key?(:owner) or options.has_key?(:group)
    raise ArgumentError, "'chown' needs at least one of owner or group"
  end
  res = operate_requests('PUT', path, 'SETOWNER', options)
  res.code == '200'
end

#content_summary(path, options = {}) ⇒ Object Also known as: getcontentsummary

curl -i “http://<HOST>:<PORT>/webhdfs/v1/<PATH>?op=GETCONTENTSUMMARY”



135
136
137
138
139
 
# File 'lib/webhdfs/client_v1.rb', line 135

def content_summary(path, options={})
  check_options(options, OPT_TABLE['GETCONTENTSUMMARY'])
  res = operate_requests('GET', path, 'GETCONTENTSUMMARY', options)
  check_success_json(res, 'ContentSummary')
end

#create(path, body, options = {}) ⇒ Object

curl -i -X PUT “http://<HOST>:<PORT>/webhdfs/v1/<PATH>?op=CREATE

[&overwrite=<true|false>][&blocksize=<LONG>][&replication=<SHORT>]
[&permission=<OCTAL>][&buffersize=<INT>]"



59
60
61
62
63
64
65
66
 
# File 'lib/webhdfs/client_v1.rb', line 59

def create(path, body, options={})
  if @httpfs_mode
    options = options.merge({'data' => 'true'})
  end
  check_options(options, OPT_TABLE['CREATE'])
  res = operate_requests('PUT', path, 'CREATE', options, body)
  res.code == '201'
end

#delete(path, options = {}) ⇒ Object

curl -i -X DELETE “http://<host>:<port>/webhdfs/v1/<path>?op=DELETE

[&recursive=<true|false>]"



111
112
113
114
115
 
# File 'lib/webhdfs/client_v1.rb', line 111

def delete(path, options={})
  check_options(options, OPT_TABLE['DELETE'])
  res = operate_requests('DELETE', path, 'DELETE', options)
  check_success_json(res, 'boolean')
end

#homedir(options = {}) ⇒ Object Also known as: gethomedirectory

curl -i “http://<HOST>:<PORT>/webhdfs/v1/?op=GETHOMEDIRECTORY”



151
152
153
154
155
 
# File 'lib/webhdfs/client_v1.rb', line 151

def homedir(options={})
  check_options(options, OPT_TABLE['GETHOMEDIRECTORY'])
  res = operate_requests('GET', '/', 'GETHOMEDIRECTORY', options)
  check_success_json(res, 'Path')
end

#list(path, options = {}) ⇒ Object Also known as: liststatus

curl -i “http://<HOST>:<PORT>/webhdfs/v1/<PATH>?op=LISTSTATUS”



127
128
129
130
131
 
# File 'lib/webhdfs/client_v1.rb', line 127

def list(path, options={})
  check_options(options, OPT_TABLE['LISTSTATUS'])
  res = operate_requests('GET', path, 'LISTSTATUS', options)
  check_success_json(res, 'FileStatuses')['FileStatus']
end

#mkdir(path, options = {}) ⇒ Object Also known as: mkdirs

curl -i -X PUT “<HOST>:<PORT>/<PATH>?op=MKDIRS”



91
92
93
94
95
 
# File 'lib/webhdfs/client_v1.rb', line 91

def mkdir(path, options={})
  check_options(options, OPT_TABLE['MKDIRS'])
  res = operate_requests('PUT', path, 'MKDIRS', options)
  check_success_json(res, 'boolean')
end

#operate_requests(method, path, op, params = {}, payload = nil) ⇒ Object 



248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
 
# File 'lib/webhdfs/client_v1.rb', line 248

def operate_requests(method, path, op, params={}, payload=nil)
  if not @httpfs_mode and REDIRECTED_OPERATIONS.include?(op)
    res = request(@host, @port, method, path, op, params, nil)
    unless res.is_a?(Net::HTTPRedirection) and res['location']
      msg = "NameNode returns non-redirection (or without location header), code:#{res.code}, body:#{res.body}."
      raise WebHDFS::RequestFailedError, msg
    end
    uri = URI.parse(res['location'])
    rpath = if uri.query
              uri.path + '?' + uri.query
            else
              uri.path
            end
    request(uri.host, uri.port, method, rpath, nil, {}, payload, {'Content-Type' => 'application/octet-stream'})
  else
    if @httpfs_mode and not payload.nil?
      request(@host, @port, method, path, op, params, payload, {'Content-Type' => 'application/octet-stream'})
    else
      request(@host, @port, method, path, op, params, payload)
    end
  end
end

#read(path, options = {}) ⇒ Object Also known as: open

curl -i -L “http://<HOST>:<PORT>/webhdfs/v1/<PATH>?op=OPEN

[&offset=<LONG>][&length=<LONG>][&buffersize=<INT>]"



82
83
84
85
86
 
# File 'lib/webhdfs/client_v1.rb', line 82

def read(path, options={})
  check_options(options, OPT_TABLE['OPEN'])
  res = operate_requests('GET', path, 'OPEN', options)
  res.body
end

#rename(path, dest, options = {}) ⇒ Object

curl -i -X PUT “<HOST>:<PORT>/webhdfs/v1/<PATH>?op=RENAME&destination=<PATH>”



100
101
102
103
104
105
106
107
 
# File 'lib/webhdfs/client_v1.rb', line 100

def rename(path, dest, options={})
  check_options(options, OPT_TABLE['RENAME'])
  unless dest.start_with?('/')
    dest = '/' + dest
  end
  res = operate_requests('PUT', path, 'RENAME', options.merge({'destination' => dest}))
  check_success_json(res, 'boolean')
end

#replication(path, replnum, options = {}) ⇒ Object Also known as: setreplication

curl -i -X PUT “http://<HOST>:<PORT>/webhdfs/v1/<PATH>?op=SETREPLICATION

[&replication=<SHORT>]"



183
184
185
186
187
 
# File 'lib/webhdfs/client_v1.rb', line 183

def replication(path, replnum, options={})
  check_options(options, OPT_TABLE['SETREPLICATION'])
  res = operate_requests('PUT', path, 'SETREPLICATION', options.merge({'replication' => replnum.to_s}))
  check_success_json(res, 'boolean')
end

#request(host, port, method, path, op = nil, params = {}, payload = nil, header = nil, retries = 0) ⇒ Object

IllegalArgumentException 400 Bad Request UnsupportedOperationException 400 Bad Request SecurityException 401 Unauthorized IOException 403 Forbidden FileNotFoundException 404 Not Found RumtimeException 500 Internal Server Error



277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
 
# File 'lib/webhdfs/client_v1.rb', line 277

def request(host, port, method, path, op=nil, params={}, payload=nil, header=nil, retries=0)
  conn = Net::HTTP.new(host, port, @proxy_address, @proxy_port)
  conn.proxy_user = @proxy_user if @proxy_user
  conn.proxy_pass = @proxy_pass if @proxy_pass
  conn.open_timeout = @open_timeout if @open_timeout
  conn.read_timeout = @read_timeout if @read_timeout

  path = Addressable::URI.escape(path) # make path safe for transmission via HTTP
  request_path = if op
                   build_path(path, op, params)
                 else
                   path
                 end
  if @ssl
    conn.use_ssl = true
    conn.ca_file = @ssl_ca_file if @ssl_ca_file
    if @ssl_verify_mode
      require 'openssl'
      conn.verify_mode = case @ssl_verify_mode
                         when :none then OpenSSL::SSL::VERIFY_NONE
                         when :peer then OpenSSL::SSL::VERIFY_PEER
                         end
    end
  end

  gsscli = nil
  if @kerberos
    require 'base64'
    require 'gssapi'
    gsscli = GSSAPI::Simple.new(@host, 'HTTP')
    token = nil
    begin
      token = gsscli.init_context
    rescue => e
      raise WebHDFS::KerberosError, e.message
    end
    if header
      header['Authorization'] = "Negotiate #{Base64.strict_encode64(token)}"
    else
      header = {'Authorization' => "Negotiate #{Base64.strict_encode64(token)}"}
    end
  end

  res = nil
  if !payload.nil? and payload.respond_to? :read and payload.respond_to? :size
    req = Net::HTTPGenericRequest.new(method,(payload ? true : false),true,request_path,header)
    raise WebHDFS::ClientError, 'Error accepting given IO resource as data payload, Not valid in methods other than PUT and POST' unless (method == 'PUT' or method == 'POST')

    req.body_stream = payload
    req.content_length = payload.size
    res = conn.request(req)
  else
    res = conn.send_request(method, request_path, payload, header)
  end

  if @kerberos and res.code == '307'
    itok = (res.header.get_fields('WWW-Authenticate') || ['']).pop.split(/\s+/).last
    unless itok
      raise WebHDFS::KerberosError, 'Server does not return WWW-Authenticate header'
    end

    begin
      gsscli.init_context(Base64.strict_decode64(itok))
    rescue => e
      raise WebHDFS::KerberosError, e.message
    end
  end

  case res
  when Net::HTTPSuccess
    res
  when Net::HTTPRedirection
    res
  else
    message = if res.body and not res.body.empty?
                res.body.gsub(/\n/, '')
              else
                'Response body is empty...'
              end

    if @retry_known_errors && retries < @retry_times
      detail = nil
      if message =~ /^\{"RemoteException":\{/
        begin
          detail = JSON.parse(message)
        rescue
          # ignore broken json response body
        end
      end
      if detail && detail['RemoteException'] && KNOWN_ERRORS.include?(detail['RemoteException']['exception'])
        sleep @retry_interval if @retry_interval > 0
        return request(host, port, method, path, op, params, payload, header, retries+1)
      end
    end

    case res.code
    when '400'
      raise WebHDFS::ClientError, message
    when '401'
      raise WebHDFS::SecurityError, message
    when '403'
      raise WebHDFS::IOError, message
    when '404'
      raise WebHDFS::FileNotFoundError, message
    when '500'
      raise WebHDFS::ServerError, message
    else
      raise WebHDFS::RequestFailedError, "response code:#{res.code}, message:#{message}"
    end
  end
end

#stat(path, options = {}) ⇒ Object Also known as: getfilestatus

curl -i “http://<HOST>:<PORT>/webhdfs/v1/<PATH>?op=GETFILESTATUS”



119
120
121
122
123
 
# File 'lib/webhdfs/client_v1.rb', line 119

def stat(path, options={})
  check_options(options, OPT_TABLE['GETFILESTATUS'])
  res = operate_requests('GET', path, 'GETFILESTATUS', options)
  check_success_json(res, 'FileStatus')
end

#touch(path, options = {}) ⇒ Object Also known as: settimes

curl -i -X PUT “http://<HOST>:<PORT>/webhdfs/v1/<PATH>?op=SETTIMES

[&modificationtime=<TIME>][&accesstime=<TIME>]"

motidicationtime: radix-10 logn integer accesstime: radix-10 logn integer



194
195
196
197
198
199
200
201
202
 
# File 'lib/webhdfs/client_v1.rb', line 194

def touch(path, options={})
  check_options(options, OPT_TABLE['SETTIMES'])
  unless options.has_key?('modificationtime') or options.has_key?('accesstime') or
      options.has_key?(:modificationtime) or options.has_key?(:accesstime)
    raise ArgumentError, "'chown' needs at least one of modificationtime or accesstime"
  end
  res = operate_requests('PUT', path, 'SETTIMES', options)
  res.code == '200'
end