v3bwfile.py 13.4 KB
Newer Older
juga's avatar
juga committed
1
2
3
4
5
# -*- coding: utf-8 -*-
"""Classes and functions that create the bandwidth measurements document
(v3bw) used by bandwidth authorities."""

import logging
juga's avatar
juga committed
6
from statistics import median
juga's avatar
juga committed
7

8
9
from sbws import __version__
from sbws.globals import SPEC_VERSION
juga's avatar
juga committed
10
from sbws.lib.resultdump import ResultSuccess, _ResultType
juga's avatar
juga committed
11
from sbws.util.filelock import FileLock
12
from sbws.util.timestamp import now_isodt_str, unixts_to_isodt_str
juga's avatar
juga committed
13
14
15

log = logging.getLogger(__name__)

16
LINE_SEP = '\n'
17
18
19
20
21
22
KEYVALUE_SEP_V110 = '='
KEYVALUE_SEP_V200 = ' '
# List of the extra KeyValues accepted by the class
EXTRA_ARG_KEYVALUES = ['software', 'software_version', 'file_created',
                       'earliest_bandwidth', 'generator_started']
# List of all unordered KeyValues currently being used to generate the file
juga's avatar
juga committed
23
UNORDERED_KEYVALUES = EXTRA_ARG_KEYVALUES + ['latest_bandwidth']
24
25
# List of all the KeyValues currently being used to generate the file
ALL_KEYVALUES = ['version'] + UNORDERED_KEYVALUES
26
TERMINATOR = '===='
juga's avatar
juga committed
27
28
# Num header lines in v1.1.0 using all the KeyValues
NUM_LINES_HEADER_V110 = len(ALL_KEYVALUES) + 2
29
30
LINE_TERMINATOR = TERMINATOR + LINE_SEP

juga's avatar
juga committed
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
# KeyValue separator in Bandwidth Lines
BW_KEYVALUE_SEP_V110 = ' '
BW_EXTRA_ARG_KEYVALUES = ['master_key_ed25519', 'nick', 'rtts', 'last_time',
                          'success', 'error_stream', 'error_circ',
                          'error_misc', 'error_auth']
BW_KEYVALUES = ['node_id', 'bw'] + BW_EXTRA_ARG_KEYVALUES


def total_bw(bw_lines):
    return sum([l.bw for l in bw_lines])


def avg_bw(bw_lines):
    assert len(bw_lines) > 0
    return total_bw(bw_lines) / len(bw_lines)


def scale_lines(bw_lines, scale_constant):
    avg = avg_bw(bw_lines)
    for line in bw_lines:
        line.bw = round(line.bw / avg * scale_constant)
    warn_if_not_accurate_enough(bw_lines, scale_constant)
    return bw_lines


def warn_if_not_accurate_enough(bw_lines, scale_constant):
    margin = 0.001
    accuracy_ratio = avg_bw(bw_lines) / scale_constant
    log.info('The generated lines are within {:.5}% of what they should '
             'be'.format((1 - accuracy_ratio) * 100))
    if accuracy_ratio < 1 - margin or accuracy_ratio > 1 + margin:
        log.warning('There was %f%% error and only +/- %f%% is '
                    'allowed', (1 - accuracy_ratio) * 100, margin * 100)

juga's avatar
juga committed
65

juga's avatar
juga committed
66
67
68
69
70
71
72
def read_started_ts(conf):
    """Read ISO formated timestamp which represents the date and time
    when scanner started.

    :param ConfigParser conf: configuration
    :returns: str, ISO formated timestamp
    """
juga's avatar
juga committed
73
74
75
76
    try:
        filepath = conf['paths']['started_filepath']
    except TypeError as e:
        return ''
juga's avatar
juga committed
77
78
79
80
81
82
83
84
85
86
    try:
        with FileLock(filepath):
            with open(filepath, 'r') as fd:
                generator_started = fd.read()
    except FileNotFoundError as e:
        log.warn('File %s not found.%s', filepath, e)
        return ''
    return generator_started


juga's avatar
juga committed
87
88
89
90
91
92
93
94
95
def num_results_of_type(results, type_str):
    return len([r for r in results if r.type == type_str])


# Better way to use enums?
def result_type_to_key(type_str):
    return type_str.replace('-', '_')


juga's avatar
juga committed
96
97
98
99
100
class V3BwHeader(object):
    """
    Create a bandwidth measurements (V3bw) header
    following bandwidth measurements document spec version 1.1.0.

101
    :param str timestamp: timestamp in Unix Epoch seconds of the most recent
102
        generator result.
juga's avatar
juga committed
103
104
105
    :param str version: the spec version
    :param str software: the name of the software that generates this
    :param str software_version: the version of the software
106
    :param dict kwargs: extra headers. Currently supported:
107
108
109
110
        - earliest_bandwidth: str, ISO 8601 timestamp in UTC time zone
          when the first bandwidth was obtained
        - generator_started: str, ISO 8601 timestamp in UTC time zone
          when the generator started
juga's avatar
juga committed
111
    """
112
    def __init__(self, timestamp, **kwargs):
juga's avatar
juga committed
113
114
115
116
        assert isinstance(timestamp, str)
        for v in kwargs.values():
            assert isinstance(v, str)
        self.timestamp = timestamp
117
118
119
120
        # KeyValues with default value when not given by kwargs
        self.version = kwargs.get('version', SPEC_VERSION)
        self.software = kwargs.get('software', 'sbws')
        self.software_version = kwargs.get('software_version', __version__)
121
        self.file_created = kwargs.get('file_created', now_isodt_str())
juga's avatar
juga committed
122
        # latest_bandwidth should not be in kwargs, since it MUST be the
123
        # same as timestamp
juga's avatar
juga committed
124
        self.latest_bandwidth = unixts_to_isodt_str(timestamp)
125
126
        [setattr(self, k, v) for k, v in kwargs.items()
         if k in EXTRA_ARG_KEYVALUES]
juga's avatar
juga committed
127

128
    @property
129
130
131
132
133
134
    def keyvalue_unordered_tuple_ls(self):
        """Return list of KeyValue tuples that do not have specific order."""
        # sort the list to generate determinist headers
        keyvalue_tuple_ls = sorted([(k, v) for k, v in self.__dict__.items()
                                    if k in UNORDERED_KEYVALUES])
        return keyvalue_tuple_ls
135
136

    @property
137
138
139
    def keyvalue_tuple_ls(self):
        """Return list of all KeyValue tuples"""
        return [('version', self.version)] + self.keyvalue_unordered_tuple_ls
140
141

    @property
142
143
144
145
146
    def keyvalue_v110str_ls(self):
        """Return KeyValue list of strings following spec v1.1.0."""
        keyvalues = [self.timestamp] + [KEYVALUE_SEP_V110.join([k, v])
                                        for k, v in self.keyvalue_tuple_ls]
        return keyvalues
147
148
149
150

    @property
    def strv110(self):
        """Return header string following spec v1.1.0."""
151
        header_str = LINE_SEP.join(self.keyvalue_v110str_ls) + LINE_SEP + \
152
153
154
155
            LINE_TERMINATOR
        return header_str

    @property
156
157
158
159
160
    def keyvalue_v200_ls(self):
        """Return KeyValue list of strings following spec v2.0.0."""
        keyvalue = [self.timestamp] + [KEYVALUE_SEP_V200.join([k, v])
                                       for k, v in self.keyvalue_tuple_ls]
        return keyvalue
161
162
163
164

    @property
    def strv200(self):
        """Return header string following spec v2.0.0."""
165
        header_str = LINE_SEP.join(self.keyvalue_v200_ls) + LINE_SEP + \
166
167
168
            LINE_TERMINATOR
        return header_str

169
170
171
172
173
    def __str__(self):
        if self.version == '1.1.0':
            return self.strv110
        return self.strv200

174
175
176
177
178
179
180
181
182
183
184
185
186
    @classmethod
    def from_lines_v110(cls, lines):
        """
        :param list lines: list of lines to parse
        :returns: tuple of V3BwHeader object and non-header lines
        """
        assert isinstance(lines, list)
        try:
            index_terminator = lines.index(TERMINATOR)
        except ValueError as e:
            # is not a bw file or is v100
            log.warn('Terminator is not in lines')
            return None
187
        ts = lines[0]
188
        kwargs = dict([l.split(KEYVALUE_SEP_V110)
189
                       for l in lines[:index_terminator]
190
                       if l.split(KEYVALUE_SEP_V110)[0] in ALL_KEYVALUES])
191
192
193
194
195
196
        h = cls(ts, **kwargs)
        return h, lines[index_terminator + 1:]

    @classmethod
    def from_text_v110(self, text):
        """
197
        :param str text: text to parse
198
199
200
201
        :returns: tuple of V3BwHeader object and non-header lines
        """
        assert isinstance(text, str)
        return self.from_lines_v110(text.split(LINE_SEP))
juga's avatar
juga committed
202
203
204
205
206
207
208
209
210
211

    @property
    def num_lines(self):
        return len(self.__str__().split(LINE_SEP))

    @staticmethod
    def generator_started_from_file(conf):
        return read_started_ts(conf)

    @staticmethod
juga's avatar
juga committed
212
    def latest_bandwidth_from_results(results):
juga's avatar
juga committed
213
        return round(max([r.time for fp in results for r in results[fp]]))
juga's avatar
juga committed
214
215
216

    @staticmethod
    def earliest_bandwidth_from_results(results):
juga's avatar
juga committed
217
        return round(min([r.time for fp in results for r in results[fp]]))
juga's avatar
juga committed
218
219
220
221

    @classmethod
    def from_results(cls, conf, results):
        kwargs = dict()
juga's avatar
juga committed
222
223
        latest_bandwidth = cls.latest_bandwidth_from_results(results)
        earliest_bandwidth = cls.latest_bandwidth_from_results(results)
juga's avatar
juga committed
224
        generator_started = cls.generator_started_from_file(conf)
juga's avatar
juga committed
225
226
        timestamp = str(latest_bandwidth)
        kwargs['latest_bandwidth'] = unixts_to_isodt_str(latest_bandwidth)
juga's avatar
juga committed
227
228
229
230
        kwargs['earliest_bandwidth'] = unixts_to_isodt_str(earliest_bandwidth)
        kwargs['generator_started'] = generator_started
        h = cls(timestamp, **kwargs)
        return h
juga's avatar
juga committed
231
232


juga's avatar
juga committed
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
class V3BWLine(object):
    def __init__(self, node_id, bw, **kwargs):
        """
        :param str node_id:
        :param int bw:
        Currently accepted KeyValues:
            - nickname, str
            - master_key_ed25519, str
            - rtt, int
            - time, str
            - sucess, int
            - error_stream, int
            - error_circ, int
            - error_misc, int
        """
        assert isinstance(node_id, str)
        assert isinstance(bw, int)
        self.node_id = node_id
        self.bw = bw
        [setattr(self, k, v) for k, v in kwargs.items()
         if k in BW_EXTRA_ARG_KEYVALUES]

    @property
    def bw_keyvalue_tuple_ls(self):
        """Return list of KeyValue Bandwidth Line tuples."""
        # sort the list to generate determinist headers
        keyvalue_tuple_ls = sorted([(k, v) for k, v in self.__dict__.items()
                                    if k in BW_KEYVALUES])
        return keyvalue_tuple_ls

    @property
    def bw_keyvalue_v110str_ls(self):
        """Return list of KeyValue Bandwidth Line strings following
        spec v1.1.0.
        """
        bw_keyvalue_str = [KEYVALUE_SEP_V110 .join([k, str(v)])
                           for k, v in self.bw_keyvalue_tuple_ls]
        return bw_keyvalue_str

    @property
    def bw_strv110(self):
        """Return Bandwidth Line string following spec v1.1.0."""
        bw_line_str = BW_KEYVALUE_SEP_V110.join(
                        self.bw_keyvalue_v110str_ls) + LINE_SEP
        return bw_line_str
juga's avatar
juga committed
278
279

    def __str__(self):
juga's avatar
juga committed
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
        return self.bw_strv110

    @classmethod
    def from_bw_line_v110(cls, line):
        assert isinstance(line, str)
        kwargs = dict([kv.split(KEYVALUE_SEP_V110)
                       for kv in line.split(BW_KEYVALUE_SEP_V110)
                       if kv.split(KEYVALUE_SEP_V110)[0] in BW_KEYVALUES])
        bw_line = cls(**kwargs)
        return bw_line

    @staticmethod
    def bw_from_results(results):
        median_bw = median([dl['amount'] / dl['duration']
                            for r in results for dl in r.downloads])
        # convert to KB and ensure it's at least 1
        bw_kb = max(round(median_bw / 1024), 1)
        return bw_kb

    @staticmethod
    def last_time_from_results(results):
        return unixts_to_isodt_str(round(max([r.time for r in results])))

    @staticmethod
    def rtts_from_results(results):
        # convert from miliseconds to seconds
        rtts = [(round(rtt * 1000)) for r in results for rtt in r.rtts]
        rtt = round(median(rtts))
        return rtt

    @staticmethod
    def result_types_from_results(results):
        rt_dict = dict([(result_type_to_key(rt.value),
                         num_results_of_type(results, rt.value))
                        for rt in _ResultType])
        return rt_dict

    @classmethod
    def from_results(cls, results):
        success_results = [r for r in results if isinstance(r, ResultSuccess)]
        log.debug('len(success_results) %s', len(success_results))
        node_id = results[0].fingerprint
        bw = cls.bw_from_results(success_results)
        kwargs = dict()
        kwargs['nick'] = results[0].nickname
        kwargs['rtt'] = cls.rtts_from_results(success_results)
        kwargs['last_time'] = cls.last_time_from_results(results)
        kwargs.update(cls.result_types_from_results(results))
        bwl = cls(node_id, bw, **kwargs)
        return bwl

    @classmethod
    def from_data(cls, data, fingerprint):
        assert fingerprint in data
        return cls.from_results(data[fingerprint])

juga's avatar
juga committed
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
388

class V3BwFile(object):
    def __init__(self, v3bwheader, v3bwlines):
        """
        :param V3BWHeader v3bwheader:
        :param list v3bwlines:
        """
        self.header = v3bwheader
        self.bw_lines = v3bwlines

    def __str__(self):
        return str(self.header) + ''.join([str(bw_line)
                                           for bw_line in self.bw_lines])

    @property
    def total_bw(self):
        return total_bw(self.bw_lines)

    @property
    def num_lines(self):
        return len(self.bw_lines)

    @property
    def avg_bw(self):
        return self.total_bw / self.num_lines

    @classmethod
    def from_results(cls, conf, output, results):
        bw_lines = [V3BWLine.from_results(results[fp]) for fp in results]
        bw_lines = sorted(bw_lines, key=lambda d: d.bw, reverse=True)
        header = V3BwHeader.from_results(conf, results)
        f = cls(header, bw_lines)
        f.write(output)
        return f

    @classmethod
    def from_arg_results(cls, args, conf, results):
        bw_lines = [V3BWLine.from_results(results[fp]) for fp in results]
        bw_lines = sorted(bw_lines, key=lambda d: d.bw, reverse=True)
        if args.scale:
            bw_lines = scale_lines(bw_lines, args.scale_constant)
        header = V3BwHeader.from_results(conf, results)
        f = cls(header, bw_lines)
        output = args.output or conf['paths']['v3bw_fname']
        f.write(output)
        return f

    def write(self, output):
        log.info('Writing v3bw file to %s', output)
        with open(output, 'wt') as fd:
            fd.write(str(self.header))
            for line in self.bw_lines:
                fd.write(str(line))