freeipa/ipaserver/install/certs.py

1263 lines
45 KiB
Python
Raw Normal View History

# Authors: Karl MacMillan <kmacmillan@mentalrootkit.com>
#
# Copyright (C) 2007 Red Hat
# see file 'COPYING' for use and warranty information
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# 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/>.
#
0000-12-31 18:09:24 -05:50
import os, stat, subprocess, re
import sys
import tempfile
import shutil
from ipapython.ipa_log_manager import *
import xml.dom.minidom
import pwd
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
import fcntl
import base64
2013-01-08 09:11:05 -06:00
from hashlib import sha1
from ConfigParser import RawConfigParser, MissingSectionHeaderError
from nss import nss
from ipapython import dogtag
from ipapython import sysrestore
from ipapython import ipautil
from ipapython import certmonger
from ipapython.certdb import get_ca_nickname
from ipapython import services as ipaservices
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
from ipapython.dn import DN
from ipalib import pkcs10, x509, api
from ipalib.errors import CertificateOperationError
from ipalib.text import _
# Apache needs access to this database so we need to create it
# where apache can reach
NSS_DIR = "/etc/httpd/alias"
CA_SERIALNO="/var/lib/ipa/ca_serialno"
def ipa_self_signed():
"""
Determine if the current IPA CA is self-signed or using another CA
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
We do this based on the CA plugin that is currently in use.
"""
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
if api.env.ra_plugin == 'selfsign':
return True
else:
return False
def ipa_self_signed_master():
"""
The selfsign backend is enabled only one a single master.
Return True/False whether this is that master.
Returns None if not a self-signed server.
"""
if ipa_self_signed():
return api.env.enable_ra
else:
return None
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
def find_cert_from_txt(cert, start=0):
"""
Given a cert blob (str) which may or may not contian leading and
trailing text, pull out just the certificate part. This will return
the FIRST cert in a stream of data.
Returns a tuple (certificate, last position in cert)
"""
s = cert.find('-----BEGIN CERTIFICATE-----', start)
e = cert.find('-----END CERTIFICATE-----', s)
if e > 0: e = e + 25
if s < 0 or e < 0:
raise RuntimeError("Unable to find certificate")
cert = cert[s:e]
return (cert, e)
def get_cert_nickname(cert):
"""
Using the subject from cert come up with a nickname suitable
for NSS. The caller can decide whether to use just the RDN
or the whole subject.
2011-07-28 13:32:26 -05:00
Returns a tuple of (rdn, subject_dn) when rdn is the string
representation of the first RDN in the subject and subject_dn
is a DN object.
"""
nsscert = x509.load_certificate(cert)
subject = str(nsscert.subject)
dn = DN(subject)
2011-07-28 13:32:26 -05:00
return (str(dn[0]), dn)
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
def next_serial(serial_file=CA_SERIALNO):
"""
Get the next serial number if we're using an NSS-based self-signed CA.
The file is an ini-like file with following properties:
lastvalue = the last serial number handed out
nextreplica = the serial number the next replica should start with
replicainterval = the number to add to nextreplica the next time a
replica is created
File locking is attempted so we have unique serial numbers.
"""
fp = None
parser = RawConfigParser()
if ipautil.file_exists(serial_file):
try:
fp = open(serial_file, "r+")
fcntl.flock(fp.fileno(), fcntl.LOCK_EX)
parser.readfp(fp)
serial = parser.getint('selfsign', 'lastvalue')
cur_serial = serial + 1
except IOError, e:
raise RuntimeError("Unable to determine serial number: %s" % str(e))
except MissingSectionHeaderError:
fcntl.flock(fp.fileno(), fcntl.LOCK_UN)
fp.close()
f=open(serial_file,"r")
r = f.readline()
f.close()
cur_serial = int(r) + 1
fp = open(serial_file, "w")
fcntl.flock(fp.fileno(), fcntl.LOCK_EX)
parser.add_section('selfsign')
parser.set('selfsign', 'nextreplica', 500000)
parser.set('selfsign', 'replicainterval', 500000)
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
else:
fp = open(serial_file, "w")
fcntl.flock(fp.fileno(), fcntl.LOCK_EX)
parser.add_section('selfsign')
parser.set('selfsign', 'nextreplica', 500000)
parser.set('selfsign', 'replicainterval', 500000)
cur_serial = 1000
try:
fp.seek(0)
parser.set('selfsign', 'lastvalue', cur_serial)
parser.write(fp)
fp.flush()
fcntl.flock(fp.fileno(), fcntl.LOCK_UN)
fp.close()
except IOError, e:
raise RuntimeError("Unable to increment serial number: %s" % str(e))
return str(cur_serial)
def next_replica(serial_file=CA_SERIALNO):
"""
Return the starting serial number for a new self-signed replica
"""
fp = None
parser = RawConfigParser()
if ipautil.file_exists(serial_file):
try:
fp = open(serial_file, "r+")
fcntl.flock(fp.fileno(), fcntl.LOCK_EX)
parser.readfp(fp)
serial = parser.getint('selfsign', 'nextreplica')
nextreplica = serial + parser.getint('selfsign', 'replicainterval')
except IOError, e:
raise RuntimeError("Unable to determine serial number: %s" % str(e))
else:
raise RuntimeError("%s does not exist, cannot create replica" % serial_file)
try:
fp.seek(0)
parser.set('selfsign', 'nextreplica', nextreplica)
parser.write(fp)
fp.flush()
fcntl.flock(fp.fileno(), fcntl.LOCK_UN)
fp.close()
except IOError, e:
raise RuntimeError("Unable to increment serial number: %s" % str(e))
return str(serial)
class NSSDatabase(object):
"""A general-purpose wrapper around a NSS cert database
For permanent NSS databases, pass the cert DB directory to __init__
For temporary databases, do not pass nssdir, and call close() when done
to remove the DB. Alternatively, a NSSDatabase can be used as a
context manager that calls close() automatically.
"""
# Traditionally, we used CertDB for our NSS DB operations, but that class
# got too tied to IPA server details, killing reusability.
# BaseCertDB is a class that knows nothing about IPA.
# Generic NSS DB code should be moved here.
def __init__(self, nssdir=None):
if nssdir is None:
self.secdir = tempfile.mkdtemp()
self._is_temporary = True
else:
self.secdir = nssdir
self._is_temporary = False
def close(self):
if self._is_temporary:
shutil.rmtree(self.secdir)
def __enter__(self):
return self
def __exit__(self, type, value, tb):
self.close()
def run_certutil(self, args, stdin=None):
new_args = ["/usr/bin/certutil", "-d", self.secdir]
new_args = new_args + args
return ipautil.run(new_args, stdin)
def create_db(self, password_filename):
"""Create cert DB
:param password_filename: Name of file containing the database password
"""
self.run_certutil(["-N", "-f", password_filename])
def list_certs(self):
"""Return nicknames and cert flags for all certs in the database
:return: List of (name, trust_flags) tuples
"""
certs, stderr, returncode = self.run_certutil(["-L"])
certs = certs.splitlines()
# FIXME, this relies on NSS never changing the formatting of certutil
certlist = []
for cert in certs:
nickname = cert[0:61]
trust = cert[61:]
if re.match(r'\w*,\w*,\w*', trust):
certlist.append((nickname.strip(), trust.strip()))
return tuple(certlist)
def find_server_certs(self):
"""Return nicknames and cert flags for server certs in the database
Server certs have an "u" character in the trust flags.
:return: List of (name, trust_flags) tuples
"""
server_certs = []
for name, flags in self.list_certs():
if 'u' in flags:
server_certs.append((name, flags))
return server_certs
def get_trust_chain(self, nickname):
"""Return names of certs in a given cert's trust chain
:param nickname: Name of the cert
:return: List of certificate names
"""
root_nicknames = []
chain, stderr, returncode = self.run_certutil([
"-O", "-n", nickname])
chain = chain.splitlines()
for c in chain:
m = re.match('\s*"(.*)" \[.*', c)
if m:
root_nicknames.append(m.groups()[0])
return root_nicknames
def import_pkcs12(self, pkcs12_filename, db_password_filename,
pkcs_password_filename=None):
args = ["/usr/bin/pk12util", "-d", self.secdir,
"-i", pkcs12_filename,
"-k", db_password_filename, '-v']
if pkcs_password_filename:
args = args + ["-w", pkcs_password_filename]
try:
ipautil.run(args)
except ipautil.CalledProcessError, e:
if e.returncode == 17:
raise RuntimeError("incorrect password for pkcs#12 file %s" %
pkcs12_filename)
else:
raise RuntimeError("unknown error import pkcs#12 file %s" %
pkcs12_filename)
def find_root_cert_from_pkcs12(self, pkcs12_fname, passwd_fname=None):
"""Given a PKCS#12 file, try to find any certificates that do
not have a key. The assumption is that these are the root CAs.
"""
args = ["/usr/bin/pk12util", "-d", self.secdir,
"-l", pkcs12_fname,
"-k", passwd_fname]
if passwd_fname:
args = args + ["-w", passwd_fname]
try:
(stdout, stderr, returncode) = ipautil.run(args)
except ipautil.CalledProcessError, e:
if e.returncode == 17:
raise RuntimeError("incorrect password for pkcs#12 file")
else:
raise RuntimeError("unknown error using pkcs#12 file")
lines = stdout.split('\n')
# A simple state machine.
# 1 = looking for a line starting with 'Certificate'
# 2 = looking for the Friendly name (nickname)
nicknames = []
state = 1
for line in lines:
if state == 2:
m = re.match("\W+Friendly Name: (.*)", line)
if m:
nicknames.append( m.groups(0)[0])
state = 1
if line == "Certificate:":
state = 2
elif not line.startswith(' '):
# Top-level item that is not a certificate
state = 1
return nicknames
def trust_root_cert(self, root_nickname):
if root_nickname[:7] == "Builtin":
root_logger.debug(
"No need to add trust for built-in root CAs, skipping %s" %
root_nickname)
else:
try:
self.run_certutil(["-M", "-n", root_nickname,
"-t", "CT,CT,"])
except ipautil.CalledProcessError, e:
raise RuntimeError(
"Setting trust on %s failed" % root_nickname)
def export_pem_cert(self, nickname, location):
"""Export the given cert to PEM file in the given location"""
cert, err, returncode = self.run_certutil(["-L", "-n", nickname, "-a"])
with open(location, "w+") as fd:
fd.write(cert)
os.chmod(location, 0444)
def import_pem_cert(self, nickname, flags, location):
"""Import a cert form the given PEM file.
The file must contain exactly one certificate.
"""
with open(location) as fd:
certs = fd.read()
cert, st = find_cert_from_txt(certs)
self.add_single_pem_cert(nickname, flags, cert)
try:
find_cert_from_txt(certs, st)
except RuntimeError:
pass
else:
raise ValueError('%s contains more than one certificate')
def add_single_pem_cert(self, nick, flags, cert):
"""Import a cert in PEM format"""
self.run_certutil(["-A", "-n", nick,
"-t", flags,
"-a"],
stdin=cert)
def verify_server_cert_validity(self, nickname, hostname):
"""Verify a certificate is valid for a SSL server with given hostname
Raises a ValueError if the certificate is invalid.
"""
certdb = cert = None
nss.nss_init(self.secdir)
try:
certdb = nss.get_default_certdb()
cert = nss.find_cert_from_nickname(nickname)
intended_usage = nss.certificateUsageSSLServer
approved_usage = cert.verify_now(certdb, True, intended_usage)
if not approved_usage & intended_usage:
raise ValueError('invalid for a SSL server')
if not cert.verify_hostname(hostname):
raise ValueError('invalid for server %s' % hostname)
finally:
del certdb, cert
nss.nss_shutdown()
return None
class CertDB(object):
"""An IPA-server-specific wrapper around NSS
This class knows IPA-specific details such as nssdir location, or the
CA cert name.
"""
# TODO: Remove all selfsign code
def __init__(self, realm, nssdir=NSS_DIR, fstore=None, host_name=None, subject_base=None):
self.nssdb = NSSDatabase(nssdir)
self.secdir = nssdir
self.realm = realm
self.noise_fname = self.secdir + "/noise.txt"
self.passwd_fname = self.secdir + "/pwdfile.txt"
self.certdb_fname = self.secdir + "/cert8.db"
self.keydb_fname = self.secdir + "/key3.db"
0000-12-31 18:09:24 -05:50
self.secmod_fname = self.secdir + "/secmod.db"
self.cacert_fname = self.secdir + "/cacert.asc"
self.pk12_fname = self.secdir + "/cacert.p12"
self.pin_fname = self.secdir + "/pin.txt"
self.pwd_conf = "/etc/httpd/conf/password.conf"
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
self.reqdir = None
self.certreq_fname = None
self.certder_fname = None
self.host_name = host_name
self.subject_base = subject_base
try:
self.cwd = os.getcwd()
except OSError, e:
raise RuntimeError("Unable to determine the current directory: %s" % str(e))
self.self_signed_ca = ipa_self_signed()
if not subject_base:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.subject_base = DN(('O', 'IPA'))
Use secure method to acquire IPA CA certificate Major changes ipa-client-install: * Use GSSAPI connection to LDAP server to download CA cert (now the default method) * Add --ca-cert-file option to load the CA cert from a disk file. Validate the file. If this option is used the supplied CA cert is considered definitive. * The insecure HTTP retrieval method is still supported but it must be explicitly forced and a warning will be emitted. * Remain backward compatible with unattended case (except for aberrant condition when preexisting /etc/ipa/ca.crt differs from securely obtained CA cert, see below) * If /etc/ipa/ca.crt CA cert preexists the validate it matches the securely acquired CA cert, if not: - If --unattended and not --force abort with error - If interactive query user to accept new CA cert, if not abort In either case warn user. * If interactive and LDAP retrieval fails prompt user if they want to proceed with insecure HTTP method * If not interactive and LDAP retrieval fails abort unless --force * Backup preexisting /etc/ipa/ca.crt in FileStore prior to execution, if ipa-client-install fails it will be restored. Other changes: * Add new exception class CertificateInvalidError * Add utility convert_ldap_error() to ipalib.ipautil * Replace all hardcoded instances of /etc/ipa/ca.crt in ipa-client-install with CACERT constant (matches existing practice elsewhere). * ipadiscovery no longer retrieves CA cert via HTTP. * Handle LDAP minssf failures during discovery, treat failure to check ldap server as a warninbg in absebce of a provided CA certificate via --ca-cert-file or though existing /etc/ipa/ca.crt file. Signed-off-by: Simo Sorce <simo@redhat.com> Signed-off-by: Rob Crittenden <rcritten@redhat.com>
2012-11-15 13:57:52 -06:00
if self.self_signed_ca:
self.cacert_name = get_ca_nickname(self.realm, 'CN=%s Certificate Authority')
else:
self.cacert_name = get_ca_nickname(self.realm)
0000-12-31 18:09:24 -05:50
self.valid_months = "120"
self.keysize = "1024"
# We are going to set the owner of all of the cert
# files to the owner of the containing directory
# instead of that of the process. This works when
# this is called by root for a daemon that runs as
# a normal user
mode = os.stat(self.secdir)
self.uid = mode[stat.ST_UID]
self.gid = mode[stat.ST_GID]
if fstore:
self.fstore = fstore
else:
self.fstore = sysrestore.FileStore('/var/lib/ipa/sysrestore')
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
subject_base = ipautil.dn_attribute_property('_subject_base')
def __del__(self):
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
if self.reqdir is not None:
shutil.rmtree(self.reqdir, ignore_errors=True)
try:
os.chdir(self.cwd)
except:
pass
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
def setup_cert_request(self):
"""
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
Create a temporary directory to store certificate requests and
certificates. This should be called before requesting certificates.
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
This is set outside of __init__ to avoid creating a temporary
directory every time we open a cert DB.
"""
if self.reqdir is not None:
return
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
self.reqdir = tempfile.mkdtemp('', 'ipa-', '/var/lib/ipa')
self.certreq_fname = self.reqdir + "/tmpcertreq"
self.certder_fname = self.reqdir + "/tmpcert.der"
# When certutil makes a request it creates a file in the cwd, make
# sure we are in a unique place when this happens
os.chdir(self.reqdir)
def set_serial_from_pkcs12(self):
"""A CA cert was loaded from a PKCS#12 file. Set up our serial file"""
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
cur_serial = self.find_cacert_serial()
try:
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
fp = open(CA_SERIALNO, "w")
parser = RawConfigParser()
parser.add_section('selfsign')
parser.set('selfsign', 'lastvalue', cur_serial)
parser.set('selfsign', 'nextreplica', 500000)
parser.set('selfsign', 'replicainterval', 500000)
parser.write(fp)
fp.close()
except IOError, e:
raise RuntimeError("Unable to increment serial number: %s" % str(e))
def set_perms(self, fname, write=False, uid=None):
if uid:
pent = pwd.getpwnam(uid)
os.chown(fname, pent.pw_uid, pent.pw_gid)
else:
os.chown(fname, self.uid, self.gid)
perms = stat.S_IRUSR
if write:
perms |= stat.S_IWUSR
os.chmod(fname, perms)
def gen_password(self):
return sha1(ipautil.ipa_generate_password()).hexdigest()
def run_certutil(self, args, stdin=None):
return self.nssdb.run_certutil(args, stdin)
def run_signtool(self, args, stdin=None):
if not self.self_signed_ca:
f = open(self.passwd_fname, "r")
password = f.readline()
f.close()
new_args = ["/usr/bin/signtool", "-d", self.secdir, "-p", password]
else:
new_args = ["/usr/bin/signtool", "-d", self.secdir]
new_args = new_args + args
ipautil.run(new_args, stdin)
def create_noise_file(self):
if ipautil.file_exists(self.noise_fname):
os.remove(self.noise_fname)
f = open(self.noise_fname, "w")
f.write(self.gen_password())
self.set_perms(self.noise_fname)
def create_passwd_file(self, passwd=None):
ipautil.backup_file(self.passwd_fname)
f = open(self.passwd_fname, "w")
if passwd is not None:
f.write("%s\n" % passwd)
else:
f.write(self.gen_password())
f.close()
self.set_perms(self.passwd_fname)
def create_certdbs(self):
ipautil.backup_file(self.certdb_fname)
ipautil.backup_file(self.keydb_fname)
0000-12-31 18:09:24 -05:50
ipautil.backup_file(self.secmod_fname)
self.nssdb.create_db(self.passwd_fname)
self.set_perms(self.passwd_fname, write=True)
def list_certs(self):
"""
Return a tuple of tuples containing (nickname, trust)
"""
return self.nssdb.list_certs()
def has_nickname(self, nickname):
"""
Returns True if nickname exists in the certdb, False otherwise.
This could also be done directly with:
certutil -L -d -n <nickname> ...
"""
certs = self.list_certs()
for cert in certs:
if nickname == cert[0]:
return True
return False
def create_ca_cert(self):
os.chdir(self.secdir)
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
subject = DN(('cn', '%s Certificate Authority' % self.realm))
p = subprocess.Popen(["/usr/bin/certutil",
"-d", self.secdir,
"-S", "-n", self.cacert_name,
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
"-s", str(subject),
"-x",
"-t", "CT,,C",
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
"-1",
"-2",
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
"-5",
"-m", next_serial(),
"-v", self.valid_months,
"-z", self.noise_fname,
"-f", self.passwd_fname],
stdin=subprocess.PIPE,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE)
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
# Create key usage extension
# 0 - Digital Signature
# 1 - Non-repudiation
# 5 - Cert signing key
# Is this a critical extension [y/N]? y
p.stdin.write("0\n1\n5\n9\ny\n")
# Create basic constraint extension
# Is this a CA certificate [y/N]? y
# Enter the path length constraint, enter to skip [<0 for unlimited pat
# Is this a critical extension [y/N]? y
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
# 5 6 7 9 n -> SSL, S/MIME, Object signing CA
p.stdin.write("y\n\ny\n")
p.stdin.write("5\n6\n7\n9\nn\n")
p.wait()
os.chdir(self.cwd)
0000-12-31 18:09:24 -05:50
def export_ca_cert(self, nickname, create_pkcs12=False):
"""create_pkcs12 tells us whether we should create a PKCS#12 file
of the CA or not. If we are running on a replica then we won't
have the private key to make a PKCS#12 file so we don't need to
do that step."""
# export the CA cert for use with other apps
ipautil.backup_file(self.cacert_fname)
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
root_nicknames = self.find_root_cert(nickname)
fd = open(self.cacert_fname, "w")
for root in root_nicknames:
(cert, stderr, returncode) = self.run_certutil(["-L", "-n", root, "-a"])
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
fd.write(cert)
fd.close()
os.chmod(self.cacert_fname, stat.S_IRUSR | stat.S_IRGRP | stat.S_IROTH)
if create_pkcs12:
ipautil.backup_file(self.pk12_fname)
ipautil.run(["/usr/bin/pk12util", "-d", self.secdir,
"-o", self.pk12_fname,
"-n", self.cacert_name,
"-w", self.passwd_fname,
"-k", self.passwd_fname])
self.set_perms(self.pk12_fname)
def load_cacert(self, cacert_fname):
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
"""
Load all the certificates from a given file. It is assumed that
this file creates CA certificates.
"""
fd = open(cacert_fname)
certs = fd.read()
fd.close()
2011-07-28 13:32:26 -05:00
ca_dn = DN(('CN','Certificate Authority'), self.subject_base)
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
st = 0
while True:
try:
(cert, st) = find_cert_from_txt(certs, st)
2011-07-28 13:32:26 -05:00
(rdn, subject_dn) = get_cert_nickname(cert)
if subject_dn == ca_dn:
nick = get_ca_nickname(self.realm)
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
else:
2011-07-28 13:32:26 -05:00
nick = str(subject_dn)
self.nssdb.add_single_pem_cert(nick, "CT,,C", cert)
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
except RuntimeError:
break
def get_cert_from_db(self, nickname, pem=True):
"""
Retrieve a certificate from the current NSS database for nickname.
pem controls whether the value returned PEM or DER-encoded. The
default is the data straight from certutil -a.
"""
try:
args = ["-L", "-n", nickname, "-a"]
(cert, err, returncode) = self.run_certutil(args)
if pem:
return cert
else:
(cert, start) = find_cert_from_txt(cert, start=0)
cert = x509.strip_header(cert)
dercert = base64.b64decode(cert)
return dercert
except ipautil.CalledProcessError:
return ''
def find_cacert_serial(self):
(out, err, returncode) = self.run_certutil(["-L", "-n", self.cacert_name])
data = out.split('\n')
for line in data:
x = re.match(r'\s+Serial Number: (\d+) .*', line)
if x is not None:
return x.group(1)
raise RuntimeError("Unable to find serial number")
def track_server_cert(self, nickname, principal, password_file=None, command=None):
"""
Tell certmonger to track the given certificate nickname.
If command is not a full path then it is prefixed with
/usr/lib[64]/ipa/certmonger.
"""
if command is not None and not os.path.isabs(command):
if sys.maxsize > 2**32:
libpath = 'lib64'
else:
libpath = 'lib'
command = '/usr/%s/ipa/certmonger/%s' % (libpath, command)
cmonger = ipaservices.knownservices.certmonger
cmonger.enable()
ipaservices.knownservices.messagebus.start()
cmonger.start()
try:
(stdout, stderr, rc) = certmonger.start_tracking(nickname, self.secdir, password_file, command)
except (ipautil.CalledProcessError, RuntimeError), e:
root_logger.error("certmonger failed starting to track certificate: %s" % str(e))
return
cmonger.stop()
cert = self.get_cert_from_db(nickname)
nsscert = x509.load_certificate(cert, dbdir=self.secdir)
subject = str(nsscert.subject)
m = re.match('New tracking request "(\d+)" added', stdout)
if not m:
root_logger.error('Didn\'t get new %s request, got %s' % (cmonger.service_name, stdout))
raise RuntimeError('%s did not issue new tracking request for \'%s\' in \'%s\'. Use \'ipa-getcert list\' to list existing certificates.' % (cmonger.service_name, nickname, self.secdir))
request_id = m.group(1)
certmonger.add_principal(request_id, principal)
certmonger.add_subject(request_id, subject)
cmonger.start()
def untrack_server_cert(self, nickname):
"""
Tell certmonger to stop tracking the given certificate nickname.
"""
# Always start certmonger. We can't untrack something if it isn't
# running
cmonger = ipaservices.knownservices.certmonger
ipaservices.knownservices.messagebus.start()
cmonger.start()
try:
certmonger.stop_tracking(self.secdir, nickname=nickname)
except (ipautil.CalledProcessError, RuntimeError), e:
root_logger.error("certmonger failed to stop tracking certificate: %s" % str(e))
cmonger.stop()
def create_server_cert(self, nickname, hostname, other_certdb=None, subject=None):
"""
other_certdb can mean one of two things, depending on the context.
If we are using a self-signed CA then other_certdb contains the
CA that will be signing our CSR.
If we are using a dogtag CA then it contains the RA agent key
that will issue our cert.
You can override the certificate Subject by specifying a subject.
Returns a certificate in DER format.
"""
0000-12-31 18:09:24 -05:50
cdb = other_certdb
if not cdb:
cdb = self
if subject is None:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
subject=DN(('CN', hostname), self.subject_base)
self.request_cert(subject)
cdb.issue_server_cert(self.certreq_fname, self.certder_fname)
self.add_cert(self.certder_fname, nickname)
fd = open(self.certder_fname, "r")
dercert = fd.read()
fd.close()
os.unlink(self.certreq_fname)
os.unlink(self.certder_fname)
return dercert
def create_signing_cert(self, nickname, hostname, other_certdb=None, subject=None):
cdb = other_certdb
if not cdb:
cdb = self
if subject is None:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
subject=DN(('CN', hostname), self.subject_base)
self.request_cert(subject)
cdb.issue_signing_cert(self.certreq_fname, self.certder_fname)
0000-12-31 18:09:24 -05:50
self.add_cert(self.certder_fname, nickname)
os.unlink(self.certreq_fname)
os.unlink(self.certder_fname)
def request_cert(self, subject, certtype="rsa", keysize="2048"):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(subject, DN)
self.create_noise_file()
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
self.setup_cert_request()
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
args = ["-R", "-s", str(subject),
"-o", self.certreq_fname,
"-k", certtype,
"-g", keysize,
"-z", self.noise_fname,
"-f", self.passwd_fname]
if not self.self_signed_ca:
args.append("-a")
(stdout, stderr, returncode) = self.run_certutil(args)
os.remove(self.noise_fname)
return (stdout, stderr)
def issue_server_cert(self, certreq_fname, cert_fname):
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
self.setup_cert_request()
if self.self_signed_ca:
p = subprocess.Popen(["/usr/bin/certutil",
"-d", self.secdir,
"-C", "-c", self.cacert_name,
"-i", certreq_fname,
"-o", cert_fname,
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
"-m", next_serial(),
"-v", self.valid_months,
"-f", self.passwd_fname,
"-1", "-5"],
stdin=subprocess.PIPE,
stdout=subprocess.PIPE)
# Bah - this sucks, but I guess it isn't possible to fully
# control this with command line arguments.
#
# What this is requesting is:
# -1 (Create key usage extension)
# 2 - Key encipherment
# 9 - done
# n - not critical
#
# -5 (Create netscape cert type extension)
# 1 - SSL Server
# 9 - done
# n - not critical
p.stdin.write("2\n9\nn\n1\n9\nn\n")
p.wait()
else:
if self.host_name is None:
raise RuntimeError("CA Host is not set.")
f = open(certreq_fname, "r")
csr = f.readlines()
f.close()
csr = "".join(csr)
# We just want the CSR bits, make sure there is nothing else
csr = pkcs10.strip_header(csr)
params = {'profileId': 'caIPAserviceCert',
'cert_request_type': 'pkcs10',
'requestor_name': 'IPA Installer',
'cert_request': csr,
'xmlOutput': 'true'}
# Send the request to the CA
f = open(self.passwd_fname, "r")
password = f.readline()
f.close()
result = dogtag.https_request(
self.host_name,
api.env.ca_ee_install_port or
dogtag.configured_constants().EE_SECURE_PORT,
"/ca/ee/ca/profileSubmitSSLClient",
self.secdir, password, "ipaCert", **params)
http_status, http_reason_phrase, http_headers, http_body = result
if http_status != 200:
raise CertificateOperationError(
error=_('Unable to communicate with CMS (%s)') %
http_reason_phrase)
# The result is an XML blob. Pull the certificate out of that
doc = xml.dom.minidom.parseString(http_body)
item_node = doc.getElementsByTagName("b64")
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
try:
try:
cert = item_node[0].childNodes[0].data
except IndexError:
raise RuntimeError("Certificate issuance failed")
finally:
doc.unlink()
# base64-decode the result for uniformity
cert = base64.b64decode(cert)
# Write the certificate to a file. It will be imported in a later
# step. This file will be read later to be imported.
f = open(cert_fname, "w")
f.write(cert)
f.close()
return
def issue_signing_cert(self, certreq_fname, cert_fname):
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
self.setup_cert_request()
if self.self_signed_ca:
p = subprocess.Popen(["/usr/bin/certutil",
"-d", self.secdir,
"-C", "-c", self.cacert_name,
"-i", certreq_fname,
"-o", cert_fname,
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
"-m", next_serial(),
"-v", self.valid_months,
"-f", self.passwd_fname,
"-1", "-5"],
stdin=subprocess.PIPE,
stdout=subprocess.PIPE)
# Bah - this sucks, but I guess it isn't possible to fully
# control this with command line arguments.
#
# What this is requesting is:
# -1 (Create key usage extension)
# 0 - Digital Signature
# 5 - Cert signing key
# 9 - done
# n - not critical
#
# -5 (Create netscape cert type extension)
# 3 - Object Signing
# 9 - done
# n - not critical
p.stdin.write("0\n5\n9\nn\n3\n9\nn\n")
p.wait()
else:
if self.host_name is None:
raise RuntimeError("CA Host is not set.")
f = open(certreq_fname, "r")
csr = f.readlines()
f.close()
csr = "".join(csr)
# We just want the CSR bits, make sure there is no thing else
csr = pkcs10.strip_header(csr)
params = {'profileId': 'caJarSigningCert',
'cert_request_type': 'pkcs10',
'requestor_name': 'IPA Installer',
'cert_request': csr,
'xmlOutput': 'true'}
# Send the request to the CA
f = open(self.passwd_fname, "r")
password = f.readline()
f.close()
result = dogtag.https_request(
self.host_name,
api.env.ca_ee_install_port or
dogtag.configured_constants().EE_SECURE_PORT,
"/ca/ee/ca/profileSubmitSSLClient",
self.secdir, password, "ipaCert", **params)
http_status, http_reason_phrase, http_headers, http_body = result
if http_status != 200:
raise RuntimeError("Unable to submit cert request")
# The result is an XML blob. Pull the certificate out of that
doc = xml.dom.minidom.parseString(http_body)
item_node = doc.getElementsByTagName("b64")
cert = item_node[0].childNodes[0].data
doc.unlink()
# base64-decode the cert for uniformity
cert = base64.b64decode(cert)
# Write the certificate to a file. It will be imported in a later
# step. This file will be read later to be imported.
f = open(cert_fname, "w")
f.write(cert)
f.close()
return
def add_cert(self, cert_fname, nickname):
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
"""
Load a certificate from a PEM file and add minimal trust.
"""
args = ["-A", "-n", nickname,
"-t", "u,u,u",
"-i", cert_fname,
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
"-f", self.passwd_fname]
self.run_certutil(args)
def create_pin_file(self):
"""
This is the format of Directory Server pin files.
"""
ipautil.backup_file(self.pin_fname)
f = open(self.pin_fname, "w")
f.write("Internal (Software) Token:")
pwdfile = open(self.passwd_fname)
f.write(pwdfile.read())
f.close()
pwdfile.close()
self.set_perms(self.pin_fname)
def create_password_conf(self):
"""
This is the format of mod_nss pin files.
"""
ipautil.backup_file(self.pwd_conf)
f = open(self.pwd_conf, "w")
f.write("internal:")
pwdfile = open(self.passwd_fname)
f.write(pwdfile.read())
f.close()
pwdfile.close()
# TODO: replace explicit uid by a platform-specific one
self.set_perms(self.pwd_conf, uid="apache")
def find_root_cert(self, nickname):
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
"""
Given a nickname, return a list of the certificates that make up
the trust chain.
"""
root_nicknames = self.nssdb.get_trust_chain(nickname)
# Try to work around a change in the F-11 certutil where untrusted
# CA's are not shown in the chain. This will make a default IPA
# server installable.
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
if len(root_nicknames) == 0 and self.self_signed_ca:
return [self.cacert_name]
0000-12-31 18:09:24 -05:50
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
return root_nicknames
def find_root_cert_from_pkcs12(self, pkcs12_fname, passwd_fname=None):
"""Given a PKCS#12 file, try to find any certificates that do
not have a key. The assumption is that these are the root CAs.
"""
args = ["/usr/bin/pk12util", "-d", self.secdir,
"-l", pkcs12_fname,
"-k", passwd_fname]
if passwd_fname:
args = args + ["-w", passwd_fname]
try:
(stdout, stderr, returncode) = ipautil.run(args)
except ipautil.CalledProcessError, e:
if e.returncode == 17:
raise RuntimeError("incorrect password")
else:
raise RuntimeError("unknown error using pkcs#12 file")
lines = stdout.split('\n')
# A simple state machine.
# 1 = looking for "Certificate:"
# 2 = looking for the Friendly name (nickname)
nicknames = []
state = 1
for line in lines:
if state == 2:
m = re.match("\W+Friendly Name: (.*)", line)
if m:
nicknames.append( m.groups(0)[0])
state = 1
if line == "Certificate:":
state = 2
return nicknames
def trust_root_cert(self, root_nickname):
if root_nickname is None:
root_logger.debug("Unable to identify root certificate to trust. Continuing but things are likely to fail.")
return
try:
self.nssdb.trust_root_cert(root_nickname)
except RuntimeError:
pass
0000-12-31 18:09:24 -05:50
def find_server_certs(self):
return self.nssdb.find_server_certs()
0000-12-31 18:09:24 -05:50
def import_pkcs12(self, pkcs12_fname, passwd_fname=None):
return self.nssdb.import_pkcs12(pkcs12_fname, self.passwd_fname,
pkcs_password_filename=passwd_fname)
0000-12-31 18:09:24 -05:50
def export_pkcs12(self, pkcs12_fname, pkcs12_pwd_fname, nickname=None):
if nickname is None:
nickname = get_ca_nickname(api.env.realm)
ipautil.run(["/usr/bin/pk12util", "-d", self.secdir,
"-o", pkcs12_fname,
"-n", nickname,
"-k", self.passwd_fname,
"-w", pkcs12_pwd_fname])
def export_pem_p12(self, pkcs12_fname, pkcs12_pwd_fname,
nickname, pem_fname):
ipautil.run(["/usr/bin/openssl", "pkcs12",
"-export", "-name", nickname,
"-in", pem_fname, "-out", pkcs12_fname,
"-passout", "file:" + pkcs12_pwd_fname])
def create_self_signed(self, passwd=None):
self.create_noise_file()
self.create_passwd_file(passwd)
self.create_certdbs()
self.create_ca_cert()
self.export_ca_cert(self.cacert_name, True)
self.create_pin_file()
def create_from_cacert(self, cacert_fname, passwd=None):
if ipautil.file_exists(self.certdb_fname):
# We already have a cert db, see if it is for the same CA.
# If it is we leave things as they are.
f = open(cacert_fname, "r")
newca = f.readlines()
f.close()
newca = "".join(newca)
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
(newca, st) = find_cert_from_txt(newca)
cacert = self.get_cert_from_db(self.cacert_name)
if cacert != '':
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
(cacert, st) = find_cert_from_txt(cacert)
if newca == cacert:
return
# The CA certificates are different or something went wrong. Start with
# a new certificate database.
self.create_passwd_file(passwd)
self.create_certdbs()
self.load_cacert(cacert_fname)
def create_from_pkcs12(self, pkcs12_fname, pkcs12_pwd_fname, passwd=None):
"""Create a new NSS database using the certificates in a PKCS#12 file.
pkcs12_fname: the filename of the PKCS#12 file
pkcs12_pwd_fname: the file containing the pin for the PKCS#12 file
nickname: the nickname/friendly-name of the cert we are loading
passwd: The password to use for the new NSS database we are creating
The global CA may be added as well in case it wasn't included in the
PKCS#12 file. Extra certs won't hurt in any case.
"""
self.create_noise_file()
self.create_passwd_file(passwd)
self.create_certdbs()
self.import_pkcs12(pkcs12_fname, pkcs12_pwd_fname)
server_certs = self.find_server_certs()
if len(server_certs) == 0:
raise RuntimeError("Could not find a suitable server cert in import in %s" % pkcs12_fname)
# We only handle one server cert
nickname = server_certs[0][0]
ca_names = [name for name, flags
in self.nssdb.list_certs() if 'u' not in flags]
if len(ca_names) == 0:
raise RuntimeError("Could not find a CA cert in %s" % pkcs12_fname)
self.cacert_name = ca_names[0]
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
for ca in ca_names:
self.trust_root_cert(ca)
self.create_pin_file()
Add external CA signing and abstract out the RA backend External CA signing is a 2-step process. You first have to run the IPA installer which will generate a CSR. You pass this CSR to your external CA and get back a cert. You then pass this cert and the CA cert and re-run the installer. The CSR is always written to /root/ipa.csr. A run would look like: # ipa-server-install --ca --external-ca -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com -U [ sign cert request ] # ipa-server-install --ca --external-ca -p password -a password --external_cert_file=/tmp/rob.crt --external_ca_file=/tmp/cacert.crt -U -p password -a password -r EXAMPLE.COM -u dirsrv -n example.com --hostname=ipa.example.com This also abstracts out the RA backend plugin so the self-signed CA we create can be used in a running server. This means that the cert plugin can request certs (and nothing else). This should let us do online replica creation. To handle the self-signed CA the simple ca_serialno file now contains additional data so we don't have overlapping serial numbers in replicas. This isn't used yet. Currently the cert plugin will not work on self-signed replicas. One very important change for self-signed CAs is that the CA is no longer held in the DS database. It is now in the Apache database. Lots of general fixes were also made in ipaserver.install.certs including: - better handling when multiple CA certificates are in a single file - A temporary directory for request certs is not always created when the class is instantiated (you have to call setup_cert_request())
2009-09-10 15:15:14 -05:00
self.export_ca_cert(nickname, False)
self.self_signed_ca=False
# This file implies that we have our own self-signed CA. Ensure
# that it no longer exists (from previous installs, for example).
try:
os.remove(CA_SERIALNO)
except:
pass
def create_kdc_cert(self, nickname, hostname, destdir):
"""Create a new certificate with the spcial othername encoding needed
by a KDC certificate.
nickname: the CN name set in the certificate
destdir: the location where cert and key are to be installed
destdir will contain kdc.pem if the operation is successful
"""
reqcfg = "kdc_req.conf"
extcfg = ipautil.SHARE_DIR + "kdc_extensions.template"
key_fname = destdir + "/kdckey.pem"
cert_fname = destdir + "/kdccert.pem"
key_cert_fname = destdir + "/kdc.pem"
# Setup the temp dir
self.setup_cert_request()
# Copy the CA password file because openssl apparently can't use
# the same file twice within the same command and throws an error
ca_pwd_file = self.reqdir + "pwdfile.txt"
shutil.copyfile(self.passwd_fname, ca_pwd_file)
# Extract the cacert.pem file used by openssl to sign the certs
ipautil.run(["/usr/bin/openssl", "pkcs12",
"-in", self.pk12_fname,
"-passin", "file:" + self.passwd_fname,
"-passout", "file:" + ca_pwd_file,
"-out", "cacert.pem"])
# Create the kdc key
ipautil.run(["/usr/bin/openssl", "genrsa",
"-out", key_fname, "2048"])
# Prepare a simple cert request
req_dict = dict(PASSWORD=self.gen_password(),
SUBJBASE=self.subject_base,
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
CERTNAME=DN(('CN', nickname)))
req_template = ipautil.SHARE_DIR + reqcfg + ".template"
conf = ipautil.template_file(req_template, req_dict)
fd = open(reqcfg, "w+")
fd.write(conf)
fd.close()
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
base = str(self.subject_base).replace(",", "/")
esc_subject = DN(('CN', '%s/%s' % (nickname, base)))
ipautil.run(["/usr/bin/openssl", "req", "-new",
"-config", reqcfg,
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
"-subj", str(esc_subject),
"-key", key_fname,
"-out", "kdc.req"])
# Finally, sign the cert using the extensions file to set the
# special name
ipautil.run(["/usr/bin/openssl", "x509", "-req",
"-CA", "cacert.pem",
"-extfile", extcfg,
"-extensions", "kdc_cert",
"-passin", "file:" + ca_pwd_file,
"-set_serial", next_serial(),
"-in", "kdc.req",
"-out", cert_fname],
env = { 'REALM':self.realm, 'HOST_FQDN':hostname })
# Merge key and cert in a single file
fd = open(key_fname, "r")
key = fd.read()
fd.close()
fd = open(cert_fname, "r")
cert = fd.read()
fd.close()
fd = open(key_cert_fname, "w")
fd.write(key)
fd.write(cert)
fd.close()
os.unlink(key_fname)
os.unlink(cert_fname)
def install_pem_from_p12(self, p12_fname, p12_pwd_fname, pem_fname):
ipautil.run(["/usr/bin/openssl", "pkcs12", "-nodes",
"-in", p12_fname, "-out", pem_fname,
"-passin", "file:" + p12_pwd_fname])
def publish_ca_cert(self, location):
shutil.copy(self.cacert_fname, location)
os.chmod(location, 0444)
def export_pem_cert(self, nickname, location):
return self.nssdb.export_pem_cert(nickname, location)