ber_scanf expects a pointer to a ber_tag_t to return the tag pointed at
by "t", if that is not provided the pointer will be store in whatever
memory location is pointed by the stack at that time causeing a crash.
Note that this is effectively unused code because in ipa-kdb the only
party that can write a key_data structure to be stored is te kdb_driver
itself and we never encode these s2kparam data.
But we need to handle this for future proofing.
Fixes#8071
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
When pulling the DM password we may have the same issues reported in
ticket #6838 for CA keys.
This commit makes sure we always check the peer has keys before any
client operation.
Ticket #6838
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Stanislav Laznicka <slaznick@redhat.com>
Reviewed-By: Michal Reznik <mreznik@redhat.com>
We need to verify the ccache is avcailable in all cases or finalize
will cause us to acquire creds with the keytab which is not what we
want.
Ticket #7037
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Stanislav Laznicka <slaznick@redhat.com>
Older clients have issues properly parsing cookies and the sessionMaxAge
setting is one of those that breaks them.
Comment out the setting and add a comment that explains why it is not
set by default.
https://pagure.io/freeipa/issue/7001
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Pavel Vomacka <pvomacka@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
In some circumstances the ccache file may disappear while
mod_auth_gssapi still has a valid cookie and the client is performing a
json server call.
This may lead to credentials getting sourced from the keytab.
Make sure we enforce what GSS NAME we want to resolve so HTTP creds are
never mistakenly sourced.
Ticket: #6972
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
In complex replication setups a replica may try to obtain CA keys from a
host that is not the master we initially create the keys against.
In this case race conditions may happen due to replication. So we need
to make sure the server we are contacting to get the CA keys has our
keys in LDAP. We do this by waiting to positively fetch our encryption
public key (the last one we create) from the target host LDAP server.
Fixes: https://pagure.io/freeipa/issue/6838
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Stanislav Laznicka <slaznick@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
When ADtrust is installed we add a PAC to all tickets, during protocol
transition we need to generate a new PAC for the requested user ticket,
not check the existing PAC on the requestor ticket.
https://pagure.io/freeipa/issue/6862
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
We slice down the received cookie so that just the content that matter
is preserved. Thi is ok because servers can't trust anything else anyway
and will accept a cookie with the ancillary data missing.
By removing variable parts like the expiry component added by
mod_session or the Expiration or Max-Age metadata we keep only the part
of the cookie that changes only when a new session is generated.
This way when storing the cookie we actually add a new entry in the
ccache only when the session actually changes, and this prevents churn
on FILE based ccaches.
Related https://pagure.io/freeipa/issue/6775
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Unfortunately the MIT krb5 library has a severe limitation with FILE
ccaches when retrieving config data. It will always only search until
the first entry is found and return that one.
For FILE caches MIT krb5 does not support removing old entries when a
new one is stored, and storage happens only in append mode, so the end
result is that even if an update is stored it is never returned with the
standard krb5_cc_get_config() call.
To work around this issue we simply implement what krb5_cc_get_config()
does under the hood with the difference that we do not stop at the first
match but keep going until all ccache entries have been checked.
Related https://pagure.io/freeipa/issue/6775
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
If cookie authentication fails and we get back a 401 see if we
tried a SPNEGO auth by checking if we had a GSSAPI context. If not
it means our session cookie was invalid or expired or some other
error happened on the server that requires us to try a full SPNEGO
handshake, so go ahead and try it.
Fixes https://pagure.io/freeipa/issue/6775
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
This new option (planned to land in gssproxy 0.7) we cache the ldap
ticket properly and avoid a ticket lookup to the KDC on each and every
ldap connection. (Also requires krb5 libs 1.15.1 to benefit from caching).
Ticket: https://pagure.io/freeipa/issue/6771
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Use the new multibase search to collect policies from multiple subtrees.
The 'any' parameter is set to 'true' so the search stop when the first result
is found in any of the bases.
https://fedorahosted.org/freeipa/ticket/6568
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Martin Babinsky <mbabinsk@redhat.com>
Internally performs multiple seraches as needed based on the basedn
strings passed in and whether the caller indicated that any result is ok
or all results are needed.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Martin Babinsky <mbabinsk@redhat.com>
GSS-SPNEGO allows us to negotiate a SASL bind with less roundtrips
therefore use it when possible.
We only enable it for local connections for now because we only
recently fixed Cyrus SASL to do proper GSS-SPNEGO negotiation. This
change means a newer and an older version are not compatible.
Restricting ourselves to the local host prevents issues with
incompatible services, and it is ok for us as we are only really
looking for speedups for the local short-lived connections performed
by the framework. Most other clients have longer lived connections,
so peformance improvements there are not as important.
Ticket: https://pagure.io/freeipa/issue/6656
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Tomas Krizek <tkrizek@redhat.com>
There were 2 issues with session logouts, one is that the logout_cookie
was checked and acted on in the wrong place, the other is that the wrong
value was set in the IPASESSION header.
Fixes https://fedorahosted.org/freeipa/ticket/6685
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Stanislav Laznicka <slaznick@redhat.com>
Removing the ccache goes too far as it will cause unrelated sessions to
fail as well, this is a problem for accounts used to do unattended
operations and that may operate in parallel.
Fixes https://fedorahosted.org/freeipa/ticket/6682
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
If we are changing identiy (different principal) insure we remove the
session cookie stored on the rpc context so that we do not mistakenly
connect with the previous identity credentials.
https://fedorahosted.org/freeipa/ticket/6543
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
Instead of relying on side effects (setting the KRB5CCNAME env var),
explicitly pass the ccache name to be used if it is not the default
ccache. This fixes some tests that sometimes fail to work properly due
to the wrong ccache being used.
https://fedorahosted.org/freeipa/ticket/6543
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
When uninstalling systemd is told to disable the service, but it is not
told to sopt it, so it believes it is still running. This can cause
issues in some cases if a reinstall is performed right after an
uninstall, as systemd may decide to stop the disabled service while we
are reinstalling, causing the new install to fail.
https://fedorahosted.org/freeipa/ticket/5959
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
Add the apache user the ipawebui group.
Make the ccaches directory owned by the ipawebui group and make
mod_auth_gssapi write the ccache files as r/w by the apache user and
the ipawebui group.
Fix tmpfiles creation ownership and permissions to allow the user to
access ccaches files.
The webui framework now works as a separate user than apache, so the certs
used to access the dogtag instance need to be usable by this new user as well.
Both apache and the webui user are in the ipawebui group, so use that.
https://fedorahosted.org/freeipa/ticket/5959
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
It seem like ALIAS_CACERT_ASC was just a redundant location for the CA
cert file which is always available in /etc/ipa/ca.crt
Just use the canonical CA cert location in /etc/ipa for all cases and
stop creating a separate cacert file.
https://fedorahosted.org/freeipa/ticket/5959
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
This is in preparation for separating out the user under which the
ipa api framework runs as.
This commit also removes certs.NSS_DIR to avoid confusion and replaces
it where appropriate with the correct NSS DB directory, either the old
HTTPD_ALIAS_DIR ot the RA DB IPA_RADB_DIR. In some cases its use is
removed altogether as it was simply not necessary.
https://fedorahosted.org/freeipa/ticket/5959
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
The anonymous user allows the framework to obtain an armor ccache without
relying on usable credentials, either via a keytab or a pkinit and
public certificates. This will be needed once the HTTP keytab is moved away
for privilege separation.
https://fedorahosted.org/freeipa/ticket/5959
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
The framework will not have direct access to the keytab anymore.
This function was used in two places, to fetch the domain list and to
re-initialize the PAC when enabling or disabling a domain trust.
The domian list is normally fetched via oddjob anyway so this use is
not necesary anymore, and the MS-PAC re-initialization can be moved
later to oddjob if needed.
https://fedorahosted.org/freeipa/ticket/5959
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
We do not want to generate runtime directories just because the packages
are installed, but only if the server is actually setup and run. Also this
will be needed later because we will create a user at install time and some
tmpfiles will need to be owned by this user.
As we are changing this code also rationalize the directory structure and
move it from the http rundir to the ipa specific rundir.
https://fedorahosted.org/freeipa/ticket/5959
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
Stop using memcache, use mod_auth_gssapi filesystem based ccaches.
Remove custom session handling, use mod_auth_gssapi and mod_session to
establish and keep a session cookie.
Add loopback to mod_auth_gssapi to do form absed auth and pass back a
valid session cookie.
And now that we do not remove ccaches files to move them to the
memcache, we can avoid the risk of pollutting the filesystem by keeping
a common ccache file for all instances of the same user.
https://fedorahosted.org/freeipa/ticket/5959
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
This is necessary to be able to successfully build archives in
environments controlled by an IPA domain which may have large uidNumbers
for user accounts.
tar-ustar allows UID/GID numbers only up to 2 million and by default a
new IPA installation can assigne UIDs in the billion range.
https://fedorahosted.org/freeipa/ticket/6418
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Petr Spacek <pspacek@redhat.com>
- Make sure to pass down the debug flag to ipa-client-install when
the server install is run in debug mode
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
- Use the correct unicode string for an error message, otherwise an
exception will generate another exception about incorrect type,
masking the original error.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The code is supposed to check that the SAN name is also authorized to be used
with the specified profile id.
The original principal has already been checked.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Fraser Tweedale <ftweedal@redhat.com>
Reviewed-By: Martin Babinsky <mbabinsk@redhat.com>
This are manual fixes for patches submitted upstream, and should be
picked up once a new asn1c is available.
They will be overridden if the code is regenerated before then.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Martin Basti <mbasti@redhat.com>
Regenerate the code with asn1c 0.9.27, this allows us to pick up a few
fixes for problems identified by coverity as well as other general bugfixes.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Martin Basti <mbasti@redhat.com>
Use a helper function to perform operations on dates in LDAP attributes.
Related to #2795
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-by: David Kupka <dkupka@redhat.com>
Like for services setting the ipaKrbAuthzData attribute on a user object will
allow us to control exactly what authz data is allowed for that user.
Setting NONE would allow no authz data, while setting MS-PAC would allow only
Active Directory compatible data.
Signed-off-by: Simo Sorce <simo@redhat.com>
Ticket: https://fedorahosted.org/freeipa/ticket/2579
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Some legacy softare is not able to properly cope with preauthentication,
allow the admins to disable the requirement to use preauthentication for
all Service Principal Names if they so desire. IPA Users are excluded,
for users, which use password of lessere entrpy, preauthentication is
always required by default.
This setting does NOT override explicit policies set on service principals
or in the global policy, it only affects the default.
Signed-off-by: Simo Sorce <simo@redhat.com>
Ticket: https://fedorahosted.org/freeipa/ticket/3860
Reviewed-By: Martin Babinsky <mbabinsk@redhat.com>
Whe requesting a keytab the salt used is the NORMAL type (for backwards and AD
compatibility), however since we added alias support we need to search for the
krbCanonicalName in preference, hen nothing is specified, and for the requested
principal name when a getkeytab operation is performed. This is so that the
correct salt can be applied. (Windows AD uses some peculiar aliases for some
special accounts to generate the salt).
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
LDAP calls just assert if an invalid (NULL) context is passed in,
so we need to be sure we have a valid connection context before
calling into LDAP APIs and fail outright if a context can't be obtained.
https://fedorahosted.org/freeipa/ticket/5577
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Martin Babinsky <mbabinsk@redhat.com>
This can be used only locally on an existing master (uses ldapi).
Useful to check the domain_level in scripts before the api is
initialized and/or credentials are available.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
move the in installutils so they can be reused by multiple scripts
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
This patch makes --setup-ca work to set upa clone CA while creating
a new replica. The standalone ipa-ca-install script is not converted
yet though.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
If the user has already run kinit try to use those credentials.
The user can always override by explicitly passing the -p flag.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
Fixes a number of places where api was not passed around internally.
Also allows to install dns in replica promotion which requires an
alternative api to be created with the right configuration.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
This patch implements a new flag --promote for the ipa-replica-install command
that allows an administrative user to 'promote' an already joined client to
become a full ipa server.
The only credentials used are that of an administrator. This code relies on
ipa-custodia being available on the peer master as well as a number of other
patches to allow a computer account to request certificates for its services.
Therefore this feature is marked to work only with domain level 1 and above
servers.
Ticket: https://fedorahosted.org/freeipa/ticket/2888
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
The DNA plugin needed to be fixed to deal with replica binddn groups.
Version 1.3.4.4 is needed for that.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
Add a customized Custodia daemon and enable it after installation.
Generates server keys and loads them in LDAP autonomously on install
or update.
Provides client code classes too.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
Instantiate CAInstall only once instead of 3 times in a row always with the
same values.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
In the dogtag/ca/kra instances self.domain is never used.
Remove it.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
If we stop or restart the server insure admin_conn gets reset or other
parts may fail to properly connect/authenticate
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
Sasl mappings can be created directly by the DS Instance, there is
no reason to create them in the krbinstance as they do not depend on
the kdc to be configured just to be created.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
Remove the custom update_key_val_in_file() and instead use the common
function config_replace_variables() available from ipautil.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Jan Cholasta <jcholast@redhat.com>
The attribute mentioned was using an older name that was later changed
in the implementation.
Also add a prominent warning about the use of the kadmin flags.
Reviewed-by: Rob Crittenden <rcritten@redhat.com>
Before krb5 1.13 the krb5_salttype_to_string() function was returning
incorrect names (display names of some kind instead of the names
used by the rest of the library to map saltname to the salt type
integer number).
This patch adds a function that checks at runtime if we have a working
function and uses a fallback map updated to the salt types known up
to 1.12, this allows us to use the library provided function in
following releases where new salt types may emerge.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-by: Milan Kubik <mkubik@redhat.com>
When kadmin tries to change a password it will get the allowed keysalts
from the password policy. Failure to provide them will result in kadmin
using the defaults specified in the kdc.conf file or hardcoded defaults
(the default salt is then of type NORMAL).
This patch provides the supported values that have been read out of the
appropriate LDAP attribute when we read the server configuration.
Then at actual password change, check if kadmin is handing us back the exact
list of supported encsalts we sent it, and in that case replace it with the
real default encsalts.
Fixes https://fedorahosted.org/freeipa/ticket/4914
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-by: Martin Babinsky <mbabinsk@redhat.com>
Since we dropped support for LANMAN hashes we do not need DES from OpenSSL
anymore. Stop including an testing for it.
Test for the MD4 algorythm instead whichis still used for the NT Hashes.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Nathaniel McCallum <npmccallum@redhat.com>
We are linking with OpenSSL in 2 files, so make it clear we intentionally
add a GPLv3 exception to allow that linking by third parties.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Nathaniel McCallum <npmccallum@redhat.com>
In this new MIT version the DAL interface changes slightly but
KRB5_KDB_DAL_MAJOR_VERSION was not changed.
Luckily KRB5_KDB_API_VERSION did change and that's enough to know
what to compile in.
Resolves: https://fedorahosted.org/freeipa/ticket/4861
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
It is not necessary to remove the ccache on upgrades on modern IPA
servers, even if the ccache contains stale data either it is re-initialized by
mod_auth_kerb or a new ccache collection is created (if completely unrelated
credentials were present), at least when using DIR or keyring ccaches.
This line causes wrong SELinux labels to be set in the kernel keyring on
uprades, which the cause the apache server to fail to use th ccache.
https://fedorahosted.org/freeipa/ticket/4815
Reviewed-By: Martin Kosek <mkosek@redhat.com>
We need to make sure we have a ld context before we can load the
configuration, otherwise ldap APIs will abort crashing the KDC.
If we have an issue connecting to LDAP the lcontext will be NULL, but
we are not checking that condition when we try to refresh the global
configuration.
https://fedorahosted.org/freeipa/ticket/4810
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Martin Kosek <mkosek@redhat.com>
This hasn't been used for a number of releases now, as ipa-kdb directly
fetches the key via LDAP.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Tomas Babej <tbabej@redhat.com>
This way make rpms will always generate new packages that can be installed on
top fo older ones, regardless of alphabetic ordering of the GIT commit id.
Also make sure version and date variables are immditely resolved, so they can't
change during the build.
Reviewed-By: Nathaniel McCallum <npmccallum@redhat.com>
A mixture of implicit and explicit tagging was being used and this caused
a bug in retrieving the enctype number due to the way ber_scanf() loosely
treat sequences and explicit tagging.
The ASN.1 notation used to describe the getkeytab operation uses implicit
tagging, so by changing the code we simply follow to the specified encoding.
Resolves: https://fedorahosted.org/freeipa/ticket/4404
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
This new extended operation is tried by default and then the code falls
back to the old method if it fails. The new method allows for server
side password generation as well as retrieval of existing credentials
w/o causing regeneration of keys on the server.
Resolves:
https://fedorahosted.org/freeipa/ticket/3859
Reviewed-By: Nathaniel McCallum <npmccallum@redhat.com>
This new extended operation allow to create new keys or retrieve
existing ones. The new set of keys is returned as a ASN.1 structure
similar to the one that is passed in by the 'set keytab' extended
operation.
Access to the operation is regulated through a new special ACI that
allows 'retrieval' only if the user has access to an attribute named
ipaProtectedOperation postfixed by the subtypes 'read_keys' and
'write_keys' to distinguish between creation and retrieval operation.
For example for allowing retrieval by a specific user the following ACI
is set on cn=accounts:
(targetattr="ipaProtectedOperation;read_keys") ...
... userattr=ipaAllowedToPerform;read_keys#USERDN)
This ACI matches only if the service object hosts a new attribute named
ipaAllowedToPerform that holds the DN of the user attempting the
operation.
Resolves:
https://fedorahosted.org/freeipa/ticket/3859
Reviewed-By: Nathaniel McCallum <npmccallum@redhat.com>
Both the password plugin and the kdb driver code automatically fall
back to the default password policy.
so stop adding an explicit reference to user objects and instead rely on the
fallback.
This way users created via the framework and users created via winsync plugin
behave the same way wrt password policies and no surprises will happen.
Also in case we need to change the default password policy DN this will allow
just code changes instead of having to change each user entry created, and
distinguish between the default policy and explicit admin changes.
Related: https://fedorahosted.org/freeipa/ticket/4085
Simo Sorce