view mercurial/sslutil.py @ 26623:5a95fe44121d

clonebundles: support for seeding clones from pre-generated bundles

Cloning can be an expensive operation for servers because the server
generates a bundle from existing repository data at request time. For
a large repository like mozilla-central, this consumes 4+ minutes
of CPU time on the server. It also results in significant network
utilization. Multiplied by hundreds or even thousands of clients and
the ensuing load can result in difficulties scaling the Mercurial server.

Despite generation of bundles being deterministic until the next
changeset is added, the generation of bundles to service a clone request
is not cached. Each clone thus performs redundant work. This is
wasteful.

This patch introduces the "clonebundles" extension and related
client-side functionality to help alleviate this deficiency. The
client-side feature is behind an experimental flag and is not enabled by
default.

It works as follows:

1) Server operator generates a bundle and makes it available on a
   server (likely HTTP).
2) Server operator defines the URL of a bundle file in a
   .hg/clonebundles.manifest file.
3) Client `hg clone`ing sees the server is advertising bundle URLs.
4) Client fetches and applies the advertised bundle.
5) Client performs equivalent of `hg pull` to fetch changes made since
   the bundle was created.

Essentially, the server performs the expensive work of generating a
bundle once and all subsequent clones fetch a static file from
somewhere. Scaling static file serving is a much more manageable
problem than scaling a Python application like Mercurial. Assuming your
repository grows less than 1% per day, the end result is 99+% of CPU
and network load from clones is eliminated, allowing Mercurial servers
to scale more easily. Serving static files also means data can be
transferred to clients as fast as they can consume it, rather than as
fast as servers can generate it. This makes clones faster.

Mozilla has implemented similar functionality of this patch on
hg.mozilla.org using a custom extension. We are hosting bundle files in
Amazon S3 and CloudFront (a CDN) and have successfully offloaded
>1 TB/day in data transfer from hg.mozilla.org, freeing up significant
bandwidth and CPU resources. The positive impact has been stellar and
I believe it has proved its value to be included in Mercurial core. I
feel it is important for the client-side support to be enabled in core
by default because it means that clients will get faster, more reliable
clones and will enable server operators to reduce load without
requiring any client-side configuration changes (assuming clients are
up to date, of course).

The scope of this feature is narrowly and specifically tailored to
cloning, despite "serve pulls from pre-generated bundles" being a valid
and useful feature. I would eventually like for Mercurial servers to
support transferring *all* repository data via statically hosted files.
You could imagine a server that siphons all pushed data to bundle files
and instructs clients to apply a stream of bundles to reconstruct all
repository data. This feature, while useful and powerful, is
significantly more work to implement because it requires the server
component have awareness of discovery and a mapping of which changesets
are in which files. Full, clone bundles, by contrast, are much simpler.

The wire protocol command is named "clonebundles" instead of something
more generic like "staticbundles" to leave the door open for a new, more
powerful and more generic server-side component with minimal backwards
compatibility implications. The name "bundleclone" is used by Mozilla's
extension and would cause problems since there are subtle differences
in Mozilla's extension.

Mozilla's experience with this idea has taught us that some form of
"content negotiation" is required. Not all clients will support all
bundle formats or even URLs (advanced TLS requirements, etc). To ensure
the highest uptake possible, a server needs to advertise multiple
versions of bundles and clients need to be able to choose the most
appropriate from that list one. The "attributes" in each
server-advertised entry facilitate this filtering and sorting. Their
use will become apparent in subsequent patches.

Initial inspiration and credit for the idea of cloning from static files
belongs to Augie Fackler and his "lookaside clone" extension proof of
concept.

# sslutil.py - SSL handling for mercurial
#
# Copyright 2005, 2006, 2007, 2008 Matt Mackall <[e-mail]>
# Copyright 2006, 2007 Alexis S. L. Carvalho <[e-mail]>
# Copyright 2006 Vadim Gelfer <[e-mail]>
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.

from __future__ import absolute_import

import os
import ssl
import sys

from .i18n import _
from . import error, util

hassni = getattr(ssl, 'HAS_SNI', False)

_canloaddefaultcerts = False
try:
    ssl_context = ssl.SSLContext
    _canloaddefaultcerts = util.safehasattr(ssl_context, 'load_default_certs')

    def wrapsocket(sock, keyfile, certfile, ui, cert_reqs=ssl.CERT_NONE,
                   ca_certs=None, serverhostname=None):
        # Allow any version of SSL starting with TLSv1 and
        # up. Note that specifying TLSv1 here prohibits use of
        # newer standards (like TLSv1_2), so this is the right way
        # to do this. Note that in the future it'd be better to
        # support using ssl.create_default_context(), which sets
        # up a bunch of things in smart ways (strong ciphers,
        # protocol versions, etc) and is upgraded by Python
        # maintainers for us, but that breaks too many things to
        # do it in a hurry.
        sslcontext = ssl.SSLContext(ssl.PROTOCOL_SSLv23)
        sslcontext.options &= ssl.OP_NO_SSLv2 & ssl.OP_NO_SSLv3
        if certfile is not None:
            def password():
                f = keyfile or certfile
                return ui.getpass(_('passphrase for %s: ') % f, '')
            sslcontext.load_cert_chain(certfile, keyfile, password)
        sslcontext.verify_mode = cert_reqs
        if ca_certs is not None:
            sslcontext.load_verify_locations(cafile=ca_certs)
        elif _canloaddefaultcerts:
            sslcontext.load_default_certs()

        sslsocket = sslcontext.wrap_socket(sock, server_hostname=serverhostname)
        # check if wrap_socket failed silently because socket had been
        # closed
        # - see http://bugs.python.org/issue13721
        if not sslsocket.cipher():
            raise error.Abort(_('ssl connection failed'))
        return sslsocket
except AttributeError:
    def wrapsocket(sock, keyfile, certfile, ui, cert_reqs=ssl.CERT_NONE,
                   ca_certs=None, serverhostname=None):
        sslsocket = ssl.wrap_socket(sock, keyfile, certfile,
                                    cert_reqs=cert_reqs, ca_certs=ca_certs,
                                    ssl_version=ssl.PROTOCOL_TLSv1)
        # check if wrap_socket failed silently because socket had been
        # closed
        # - see http://bugs.python.org/issue13721
        if not sslsocket.cipher():
            raise error.Abort(_('ssl connection failed'))
        return sslsocket

def _verifycert(cert, hostname):
    '''Verify that cert (in socket.getpeercert() format) matches hostname.
    CRLs is not handled.

    Returns error message if any problems are found and None on success.
    '''
    if not cert:
        return _('no certificate received')
    dnsname = hostname.lower()
    def matchdnsname(certname):
        return (certname == dnsname or
                '.' in dnsname and certname == '*.' + dnsname.split('.', 1)[1])

    san = cert.get('subjectAltName', [])
    if san:
        certnames = [value.lower() for key, value in san if key == 'DNS']
        for name in certnames:
            if matchdnsname(name):
                return None
        if certnames:
            return _('certificate is for %s') % ', '.join(certnames)

    # subject is only checked when subjectAltName is empty
    for s in cert.get('subject', []):
        key, value = s[0]
        if key == 'commonName':
            try:
                # 'subject' entries are unicode
                certname = value.lower().encode('ascii')
            except UnicodeEncodeError:
                return _('IDN in certificate not supported')
            if matchdnsname(certname):
                return None
            return _('certificate is for %s') % certname
    return _('no commonName or subjectAltName found in certificate')


# CERT_REQUIRED means fetch the cert from the server all the time AND
# validate it against the CA store provided in web.cacerts.

def _plainapplepython():
    """return true if this seems to be a pure Apple Python that
    * is unfrozen and presumably has the whole mercurial module in the file
      system
    * presumably is an Apple Python that uses Apple OpenSSL which has patches
      for using system certificate store CAs in addition to the provided
      cacerts file
    """
    if sys.platform != 'darwin' or util.mainfrozen() or not sys.executable:
        return False
    exe = os.path.realpath(sys.executable).lower()
    return (exe.startswith('/usr/bin/python') or
            exe.startswith('/system/library/frameworks/python.framework/'))

def _defaultcacerts():
    """return path to CA certificates; None for system's store; ! to disable"""
    if _plainapplepython():
        dummycert = os.path.join(os.path.dirname(__file__), 'dummycert.pem')
        if os.path.exists(dummycert):
            return dummycert
    if _canloaddefaultcerts:
        return None
    return '!'

def sslkwargs(ui, host):
    kws = {'ui': ui}
    hostfingerprint = ui.config('hostfingerprints', host)
    if hostfingerprint:
        return kws
    cacerts = ui.config('web', 'cacerts')
    if cacerts == '!':
        pass
    elif cacerts:
        cacerts = util.expandpath(cacerts)
        if not os.path.exists(cacerts):
            raise error.Abort(_('could not find web.cacerts: %s') % cacerts)
    else:
        cacerts = _defaultcacerts()
        if cacerts and cacerts != '!':
            ui.debug('using %s to enable OS X system CA\n' % cacerts)
        ui.setconfig('web', 'cacerts', cacerts, 'defaultcacerts')
    if cacerts != '!':
        kws.update({'ca_certs': cacerts,
                    'cert_reqs': ssl.CERT_REQUIRED,
                    })
    return kws

class validator(object):
    def __init__(self, ui, host):
        self.ui = ui
        self.host = host

    def __call__(self, sock, strict=False):
        host = self.host
        cacerts = self.ui.config('web', 'cacerts')
        hostfingerprint = self.ui.config('hostfingerprints', host)

        if not sock.cipher(): # work around http://bugs.python.org/issue13721
            raise error.Abort(_('%s ssl connection error') % host)
        try:
            peercert = sock.getpeercert(True)
            peercert2 = sock.getpeercert()
        except AttributeError:
            raise error.Abort(_('%s ssl connection error') % host)

        if not peercert:
            raise error.Abort(_('%s certificate error: '
                               'no certificate received') % host)
        peerfingerprint = util.sha1(peercert).hexdigest()
        nicefingerprint = ":".join([peerfingerprint[x:x + 2]
            for x in xrange(0, len(peerfingerprint), 2)])
        if hostfingerprint:
            if peerfingerprint.lower() != \
                    hostfingerprint.replace(':', '').lower():
                raise error.Abort(_('certificate for %s has unexpected '
                                   'fingerprint %s') % (host, nicefingerprint),
                                 hint=_('check hostfingerprint configuration'))
            self.ui.debug('%s certificate matched fingerprint %s\n' %
                          (host, nicefingerprint))
        elif cacerts != '!':
            msg = _verifycert(peercert2, host)
            if msg:
                raise error.Abort(_('%s certificate error: %s') % (host, msg),
                                 hint=_('configure hostfingerprint %s or use '
                                        '--insecure to connect insecurely') %
                                      nicefingerprint)
            self.ui.debug('%s certificate successfully verified\n' % host)
        elif strict:
            raise error.Abort(_('%s certificate with fingerprint %s not '
                               'verified') % (host, nicefingerprint),
                             hint=_('check hostfingerprints or web.cacerts '
                                     'config setting'))
        else:
            self.ui.warn(_('warning: %s certificate with fingerprint %s not '
                           'verified (check hostfingerprints or web.cacerts '
                           'config setting)\n') %
                         (host, nicefingerprint))