url_helper.py 6.8 KB
Newer Older
Joshua Harlow's avatar
Joshua Harlow committed
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
# 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/>.

23
from contextlib import closing
Joshua Harlow's avatar
Joshua Harlow committed
24

25
import errno
26
import socket
27 28 29 30 31
import time
import urllib
import urllib2

from cloudinit import log as logging
32
from cloudinit import version
33 34 35 36

LOG = logging.getLogger(__name__)


37 38 39 40 41
def ok_http_code(st, redirects_ok=False):
    if redirects_ok:
        return st in xrange(200, 400)
    else:
        return st in xrange(200, 300)
42 43


44 45
def readurl(url, data=None, timeout=None,
            retries=0, sec_between=1, headers=None):
46

47 48 49 50
    req_args = {}
    req_args['url'] = url
    if data is not None:
        req_args['data'] = urllib.urlencode(data)
51 52 53 54 55 56 57

    if not headers:
        headers = {
            'User-Agent': 'Cloud-Init/%s' % (version.version_string()),
        }

    req_args['headers'] = headers
58
    req = urllib2.Request(**req_args)
59

60
    retries = max(retries, 0)
61
    attempts = retries + 1
62

63
    excepts = []
64 65 66 67 68
    LOG.info(("Attempting to read from %s with %s attempts"
                " (%s retries) to be performed"), url, attempts, retries)
    open_args = {}
    if timeout is not None:
        open_args['timeout'] = int(timeout)
69
    for i in range(0, attempts):
70
        try:
71 72 73 74 75 76 77 78 79
            with closing(urllib2.urlopen(req, **open_args)) as rh:
                content = rh.read()
                status = rh.getcode()
                if status is None:
                    # This seems to happen when files are read...
                    status = 200
                LOG.info("Read from %s (%s, %sb) after %s attempts",
                         url, status, len(content), (i + 1))
                return (content, status)
80
        except urllib2.HTTPError as e:
81
            excepts.append(e)
82 83
        except urllib2.URLError as e:
            # This can be a message string or
84 85
            # another exception instance 
            # (socket.error for remote URLs, OSError for local URLs).
86
            if (isinstance(e.reason, (OSError)) and
87
                e.reason.errno == errno.ENOENT):
88
                excepts.append(e.reason)
89
            else:
90
                excepts.append(e)
91
        if i + 1 < attempts:
92
            LOG.debug("Please wait %s seconds while we wait to try again",
93
                     sec_between)
94
            time.sleep(sec_between)
95 96

    # Didn't work out
97
    LOG.warn("Failed reading from %s after %s attempts", url, attempts)
98 99 100 101 102
    
    # It must of errored at least once for code
    # to get here so re-raise the last error
    LOG.debug("%s errors occured, re-raising the last one", len(excepts))
    raise excepts[-1]
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


def wait_for_url(urls, max_wait=None, timeout=None,
                 status_cb=None, headers_cb=None, sleep_time=1):
    """
    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.
    timeout:   the timeout provided to urllib2.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.

    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.
    """
132
    start_time = time.time()
133

134 135
    def log_status_cb(msg):
        LOG.info(msg)
136 137

    if status_cb is None:
138
        status_cb = log_status_cb
139

140
    def timeup(max_wait, start_time):
141
        return ((max_wait <= 0 or max_wait is None) or
142
                (time.time() - start_time > max_wait))
143 144 145

    loop_n = 0
    while True:
146
        sleep_time = int(loop_n / 5) + 1
147 148 149
        for url in urls:
            now = time.time()
            if loop_n != 0:
150
                if timeup(max_wait, start_time):
151
                    break
152
                if timeout and (now + timeout > (start_time + max_wait)):
153
                    # shorten timeout to not run way over max_time
154
                    timeout = int((start_time + max_wait) - now)
155 156 157 158 159 160 161 162

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

163
                (resp, sc) = readurl(url, headers=headers, timeout=timeout)
164
                if not resp:
165 166 167
                    reason = "empty response [%s]" % sc
                elif not ok_http_code(sc):
                    reason = "bad status code [%s]" % sc
168 169 170 171 172 173 174 175 176 177 178
                else:
                    return url
            except urllib2.HTTPError as e:
                reason = "http error [%s]" % e.code
            except urllib2.URLError as e:
                reason = "url error [%s]" % e.reason
            except socket.timeout as e:
                reason = "socket timeout [%s]" % e
            except Exception as e:
                reason = "unexpected error [%s]" % e

179 180 181 182 183
            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)
184

185
        if timeup(max_wait, start_time):
186 187 188
            break

        loop_n = loop_n + 1
189 190 191
        LOG.info("Please wait %s seconds while we wait to try again",
                 sleep_time)
        time.sleep(sleep_time)
192 193

    return False