~oddbloke/cloud-init/lp1506187

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
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
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
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
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
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
# vi: ts=4 expandtab
#
#    Copyright (C) 2012 Canonical Ltd.
#    Copyright (C) 2012 Hewlett-Packard Development Company, L.P.
#    Copyright (C) 2012 Yahoo! Inc.
#
#    Author: Scott Moser <scott.moser@canonical.com>
#    Author: Juerg Haefliger <juerg.haefliger@hp.com>
#    Author: Joshua Harlow <harlowja@yahoo-inc.com>
#
#    This program is free software: you can redistribute it and/or modify
#    it under the terms of the GNU General Public License version 3, as
#    published by the Free Software Foundation.
#
#    This program is distributed in the hope that it will be useful,
#    but WITHOUT ANY WARRANTY; without even the implied warranty of
#    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
#    GNU General Public License for more details.
#
#    You should have received a copy of the GNU General Public License
#    along with this program.  If not, see <http://www.gnu.org/licenses/>.

import json
import os
import requests
import six
import time

from email.utils import parsedate
from functools import partial
from requests import exceptions
import oauthlib.oauth1 as oauth1

from six.moves.urllib.parse import (
    urlparse, urlunparse,
    quote as urlquote)

from cloudinit import log as logging
from cloudinit import version

LOG = logging.getLogger(__name__)

if six.PY2:
    import httplib
    NOT_FOUND = httplib.NOT_FOUND
else:
    import http.client
    NOT_FOUND = http.client.NOT_FOUND


# Check if requests has ssl support (added in requests >= 0.8.8)
SSL_ENABLED = False
CONFIG_ENABLED = False  # This was added in 0.7 (but taken out in >=1.0)
_REQ_VER = None
try:
    from distutils.version import LooseVersion
    import pkg_resources
    _REQ = pkg_resources.get_distribution('requests')
    _REQ_VER = LooseVersion(_REQ.version)
    if _REQ_VER >= LooseVersion('0.8.8'):
        SSL_ENABLED = True
    if _REQ_VER >= LooseVersion('0.7.0') and _REQ_VER < LooseVersion('1.0.0'):
        CONFIG_ENABLED = True
except:
    pass


def _cleanurl(url):
    parsed_url = list(urlparse(url, scheme='http'))
    if not parsed_url[1] and parsed_url[2]:
        # Swap these since this seems to be a common
        # occurrence when given urls like 'www.google.com'
        parsed_url[1] = parsed_url[2]
        parsed_url[2] = ''
    return urlunparse(parsed_url)


def combine_url(base, *add_ons):

    def combine_single(url, add_on):
        url_parsed = list(urlparse(url))
        path = url_parsed[2]
        if path and not path.endswith("/"):
            path += "/"
        path += urlquote(str(add_on), safe="/:")
        url_parsed[2] = path
        return urlunparse(url_parsed)

    url = base
    for add_on in add_ons:
        url = combine_single(url, add_on)
    return url


# Made to have same accessors as UrlResponse so that the
# read_file_or_url can return this or that object and the
# 'user' of those objects will not need to know the difference.
class StringResponse(object):
    def __init__(self, contents, code=200):
        self.code = code
        self.headers = {}
        self.contents = contents
        self.url = None

    def ok(self, *args, **kwargs):
        if self.code != 200:
            return False
        return True

    def __str__(self):
        return self.contents


class FileResponse(StringResponse):
    def __init__(self, path, contents, code=200):
        StringResponse.__init__(self, contents, code=code)
        self.url = path


class UrlResponse(object):
    def __init__(self, response):
        self._response = response

    @property
    def contents(self):
        return self._response.content

    @property
    def url(self):
        return self._response.url

    def ok(self, redirects_ok=False):
        upper = 300
        if redirects_ok:
            upper = 400
        if self.code >= 200 and self.code < upper:
            return True
        else:
            return False

    @property
    def headers(self):
        return self._response.headers

    @property
    def code(self):
        return self._response.status_code

    def __str__(self):
        return self._response.text


class UrlError(IOError):
    def __init__(self, cause, code=None, headers=None, url=None):
        IOError.__init__(self, str(cause))
        self.cause = cause
        self.code = code
        self.headers = headers
        if self.headers is None:
            self.headers = {}
        self.url = url


def _get_ssl_args(url, ssl_details):
    ssl_args = {}
    scheme = urlparse(url).scheme
    if scheme == 'https' and ssl_details:
        if not SSL_ENABLED:
            LOG.warn("SSL is not supported in requests v%s, "
                     "cert. verification can not occur!", _REQ_VER)
        else:
            if 'ca_certs' in ssl_details and ssl_details['ca_certs']:
                ssl_args['verify'] = ssl_details['ca_certs']
            else:
                ssl_args['verify'] = True
            if 'cert_file' in ssl_details and 'key_file' in ssl_details:
                ssl_args['cert'] = [ssl_details['cert_file'],
                                    ssl_details['key_file']]
            elif 'cert_file' in ssl_details:
                ssl_args['cert'] = str(ssl_details['cert_file'])
    return ssl_args


def readurl(url, data=None, timeout=None, retries=0, sec_between=1,
            headers=None, headers_cb=None, ssl_details=None,
            check_status=True, allow_redirects=True, exception_cb=None):
    url = _cleanurl(url)
    req_args = {
        'url': url,
    }
    req_args.update(_get_ssl_args(url, ssl_details))
    req_args['allow_redirects'] = allow_redirects
    req_args['method'] = 'GET'
    if timeout is not None:
        req_args['timeout'] = max(float(timeout), 0)
    if data:
        req_args['method'] = 'POST'
    # It doesn't seem like config
    # was added in older library versions (or newer ones either), thus we
    # need to manually do the retries if it wasn't...
    if CONFIG_ENABLED:
        req_config = {
            'store_cookies': False,
        }
        # Don't use the retry support built-in
        # since it doesn't allow for 'sleep_times'
        # in between tries....
        # if retries:
        #     req_config['max_retries'] = max(int(retries), 0)
        req_args['config'] = req_config
    manual_tries = 1
    if retries:
        manual_tries = max(int(retries) + 1, 1)

    def_headers = {
        'User-Agent': 'Cloud-Init/%s' % (version.version_string()),
    }
    if headers:
        def_headers.update(headers)
    headers = def_headers

    if not headers_cb:
        def _cb(url):
            return headers
        headers_cb = _cb
    if data:
        req_args['data'] = data
    if sec_between is None:
        sec_between = -1

    excps = []
    # Handle retrying ourselves since the built-in support
    # doesn't handle sleeping between tries...
    for i in range(0, manual_tries):
        req_args['headers'] = headers_cb(url)
        filtered_req_args = {}
        for (k, v) in req_args.items():
            if k == 'data':
                continue
            filtered_req_args[k] = v
        try:
            LOG.debug("[%s/%s] open '%s' with %s configuration", i,
                      manual_tries, url, filtered_req_args)

            r = requests.request(**req_args)
            if check_status:
                r.raise_for_status()
            LOG.debug("Read from %s (%s, %sb) after %s attempts", url,
                      r.status_code, len(r.content), (i + 1))
            # Doesn't seem like we can make it use a different
            # subclass for responses, so add our own backward-compat
            # attrs
            return UrlResponse(r)
        except exceptions.RequestException as e:
            if (isinstance(e, (exceptions.HTTPError))
                    and hasattr(e, 'response')  # This appeared in v 0.10.8
                    and hasattr(e.response, 'status_code')):
                excps.append(UrlError(e, code=e.response.status_code,
                                      headers=e.response.headers,
                                      url=url))
            else:
                excps.append(UrlError(e, url=url))
                if SSL_ENABLED and isinstance(e, exceptions.SSLError):
                    # ssl exceptions are not going to get fixed by waiting a
                    # few seconds
                    break
            if exception_cb and exception_cb(req_args.copy(), excps[-1]):
                # if an exception callback was given it should return None
                # a true-ish value means to break and re-raise the exception
                break
            if i + 1 < manual_tries and sec_between > 0:
                LOG.debug("Please wait %s seconds while we wait to try again",
                          sec_between)
                time.sleep(sec_between)
    if excps:
        raise excps[-1]
    return None  # Should throw before this...


def wait_for_url(urls, max_wait=None, timeout=None,
                 status_cb=None, headers_cb=None, sleep_time=1,
                 exception_cb=None):
    """
    urls:      a list of urls to try
    max_wait:  roughly the maximum time to wait before giving up
               The max time is *actually* len(urls)*timeout as each url will
               be tried once and given the timeout provided.
               a number <= 0 will always result in only one try
    timeout:   the timeout provided to urlopen
    status_cb: call method with string message when a url is not available
    headers_cb: call method with single argument of url to get headers
                for request.
    exception_cb: call method with 2 arguments 'msg' (per status_cb) and
                  'exception', the exception that occurred.

    the idea of this routine is to wait for the EC2 metdata service to
    come up.  On both Eucalyptus and EC2 we have seen the case where
    the instance hit the MD before the MD service was up.  EC2 seems
    to have permenantely fixed this, though.

    In openstack, the metadata service might be painfully slow, and
    unable to avoid hitting a timeout of even up to 10 seconds or more
    (LP: #894279) for a simple GET.

    Offset those needs with the need to not hang forever (and block boot)
    on a system where cloud-init is configured to look for EC2 Metadata
    service but is not going to find one.  It is possible that the instance
    data host (169.254.169.254) may be firewalled off Entirely for a sytem,
    meaning that the connection will block forever unless a timeout is set.
    """
    start_time = time.time()

    def log_status_cb(msg, exc=None):
        LOG.debug(msg)

    if status_cb is None:
        status_cb = log_status_cb

    def timeup(max_wait, start_time):
        return ((max_wait <= 0 or max_wait is None) or
                (time.time() - start_time > max_wait))

    loop_n = 0
    while True:
        sleep_time = int(loop_n / 5) + 1
        for url in urls:
            now = time.time()
            if loop_n != 0:
                if timeup(max_wait, start_time):
                    break
                if timeout and (now + timeout > (start_time + max_wait)):
                    # shorten timeout to not run way over max_time
                    timeout = int((start_time + max_wait) - now)

            reason = ""
            url_exc = None
            try:
                if headers_cb is not None:
                    headers = headers_cb(url)
                else:
                    headers = {}

                response = readurl(url, headers=headers, timeout=timeout,
                                   check_status=False)
                if not response.contents:
                    reason = "empty response [%s]" % (response.code)
                    url_exc = UrlError(ValueError(reason), code=response.code,
                                       headers=response.headers, url=url)
                elif not response.ok():
                    reason = "bad status code [%s]" % (response.code)
                    url_exc = UrlError(ValueError(reason), code=response.code,
                                       headers=response.headers, url=url)
                else:
                    return url
            except UrlError as e:
                reason = "request error [%s]" % e
                url_exc = e
            except Exception as e:
                reason = "unexpected error [%s]" % e
                url_exc = e

            time_taken = int(time.time() - start_time)
            status_msg = "Calling '%s' failed [%s/%ss]: %s" % (url,
                                                               time_taken,
                                                               max_wait,
                                                               reason)
            status_cb(status_msg)
            if exception_cb:
                # This can be used to alter the headers that will be sent
                # in the future, for example this is what the MAAS datasource
                # does.
                exception_cb(msg=status_msg, exception=url_exc)

        if timeup(max_wait, start_time):
            break

        loop_n = loop_n + 1
        LOG.debug("Please wait %s seconds while we wait to try again",
                  sleep_time)
        time.sleep(sleep_time)

    return False


class OauthUrlHelper(object):
    def __init__(self, consumer_key=None, token_key=None,
                 token_secret=None, consumer_secret=None,
                 skew_data_file="/run/oauth_skew.json"):
        self.consumer_key = consumer_key
        self.consumer_secret = consumer_secret or ""
        self.token_key = token_key
        self.token_secret = token_secret
        self.skew_data_file = skew_data_file
        self._do_oauth = True
        self.skew_change_limit = 5
        required = (self.token_key, self.token_secret, self.consumer_key)
        if not any(required):
            self._do_oauth = False
        elif not all(required):
            raise ValueError("all or none of token_key, token_secret, or "
                             "consumer_key can be set")

        old = self.read_skew_file()
        self.skew_data = old or {}

    def read_skew_file(self):
        if self.skew_data_file and os.path.isfile(self.skew_data_file):
            with open(self.skew_data_file, mode="r") as fp:
                return json.load(fp)
        return None

    def update_skew_file(self, host, value):
        # this is not atomic
        if not self.skew_data_file:
            return
        cur = self.read_skew_file()
        if cur is None:
            cur = {}
        cur[host] = value
        with open(self.skew_data_file, mode="w") as fp:
            fp.write(json.dumps(cur))

    def exception_cb(self, msg, exception):
        if not (isinstance(exception, UrlError) and
                (exception.code == 403 or exception.code == 401)):
            return

        if 'date' not in exception.headers:
            LOG.warn("Missing header 'date' in %s response", exception.code)
            return

        date = exception.headers['date']
        try:
            remote_time = time.mktime(parsedate(date))
        except Exception as e:
            LOG.warn("Failed to convert datetime '%s': %s", date, e)
            return

        skew = int(remote_time - time.time())
        host = urlparse(exception.url).netloc
        old_skew = self.skew_data.get(host, 0)
        if abs(old_skew - skew) > self.skew_change_limit:
            self.update_skew_file(host, skew)
            LOG.warn("Setting oauth clockskew for %s to %d", host, skew)
        self.skew_data[host] = skew

        return

    def headers_cb(self, url):
        if not self._do_oauth:
            return {}

        timestamp = None
        host = urlparse(url).netloc
        if self.skew_data and host in self.skew_data:
            timestamp = int(time.time()) + self.skew_data[host]

        return oauth_headers(
            url=url, consumer_key=self.consumer_key,
            token_key=self.token_key, token_secret=self.token_secret,
            consumer_secret=self.consumer_secret, timestamp=timestamp)

    def _wrapped(self, wrapped_func, args, kwargs):
        kwargs['headers_cb'] = partial(
            self._headers_cb, kwargs.get('headers_cb'))
        kwargs['exception_cb'] = partial(
            self._exception_cb, kwargs.get('exception_cb'))
        return wrapped_func(*args, **kwargs)

    def wait_for_url(self, *args, **kwargs):
        return self._wrapped(wait_for_url, args, kwargs)

    def readurl(self, *args, **kwargs):
        return self._wrapped(readurl, args, kwargs)

    def _exception_cb(self, extra_exception_cb, msg, exception):
        ret = None
        try:
            if extra_exception_cb:
                ret = extra_exception_cb(msg, exception)
        finally:
                self.exception_cb(msg, exception)
        return ret

    def _headers_cb(self, extra_headers_cb, url):
        headers = {}
        if extra_headers_cb:
            headers = extra_headers_cb(url)
        headers.update(self.headers_cb(url))
        return headers


def oauth_headers(url, consumer_key, token_key, token_secret, consumer_secret,
                  timestamp=None):
    if timestamp:
        timestamp = str(timestamp)
    else:
        timestamp = None

    client = oauth1.Client(
        consumer_key,
        client_secret=consumer_secret,
        resource_owner_key=token_key,
        resource_owner_secret=token_secret,
        signature_method=oauth1.SIGNATURE_PLAINTEXT,
        timestamp=timestamp)
    uri, signed_headers, body = client.sign(url)
    return signed_headers