Age | Commit message (Collapse) | Author | Files | Lines |
|
to DRS
This is a very serious situation (it should not happen) so getting information on the
objects that this happens for would be very helpful.
Andrew Bartlett
Reviewed-by: Stefan Metzmacher <metze@samba.org>
|
|
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
|
|
This needs to be more async, and give less scary errors.
Andrew Bartlett
Reviewed-by: Stefan Metzmacher <metze@samba.org>
|
|
We don't behave completely like a Windows server, but it's much more
identical than before.
The partition head is always the first object followed by the rest
sorted by uSNChanged.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Tue Jan 1 21:09:42 CET 2013 on sn-devel-104
|
|
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
|
|
We should build the final highwatermark and uptodatevector of
a replication cycle at the start of the cycle. Before we
search for the currently missing objects.
Otherwise we risk that some objects get lost.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
|
|
The given highwatermark is only valid relative to the
specified source_dsa_invocation_id.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
|
|
replication cycle (bug #9508)
If the highwatermark given by the client is not the one we expect,
we need to start a new replication cycle. Otherwise the destination dsa
skips objects and linked attribute values.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
|
|
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
|
|
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
|
|
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
|
|
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
|
|
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
|
|
|
|
Check that both the GUID and DN are the GUID/DN of a NC if not return
WERR_DS_DRA_BAD_NC
|
|
|
|
|
|
Paragraph 4.1.10.5 says that
if err = 0 then
msgOut.pNC := msgIn.pNC
msgOut.usnvecFrom := msgIn.usnvecFrom
so no need to set the highestUsn to 0
|
|
|
|
|
|
pointing to a DC object
The problem was found by the DRSR testsuite where server objects were
created in the Site container without serverrefrence attribute
triggering error in the testsuite.
|
|
dcesrv_drsuapi_DsBind()
metze
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Tue Sep 25 03:06:13 CEST 2012 on sn-devel-104
|
|
dcesrv_drsuapi_DsGetDomainControllerInfo_1()
metze
|
|
|
|
As this value is calculated new each time, we need to give it a context to live on.
If the value is the forced value during provision, a reference is taken.
This was responsible for the memory leak in the replication process. In the
example I was given, this DN appeared in memory 13596 times!
Andrew Bartlett
Autobuild-User(master): Andrew Bartlett <abartlet@samba.org>
Autobuild-Date(master): Tue Aug 14 10:05:14 CEST 2012 on sn-devel-104
|
|
|
|
metze
Autobuild-User: Matthieu Patou <mat@samba.org>
Autobuild-Date: Fri Dec 23 09:30:09 CET 2011 on sn-devel-104
|
|
Signed-off-by: Stefan Metzmacher <metze@samba.org>
|
|
metze
|
|
something in less than x seconds
Signed-off-by: Andrew Tridgell <tridge@samba.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
|
|
unefficient
Current implementation of ldb_msg_find_attr_as_* iterate on the list of
attributes returned by the search and make a string comparison. As we
sorting the array of messages / guids we tend to call this function many
times. By storing the GUID and the USN in a separate structure we are
sure to call this function only once per attribute and object.
Signed-off-by: Andrew Tridgell <tridge@samba.org>
|
|
not exists and so is always False
Having a false value cause the link on removed attribute to be always
returned which is what we try to avoid.
|
|
we want new NCs to be created
|
|
when a DC has the GUID_DRS_GET_FILTERED_ATTRIBUTES right on a NC, we
need to allow it to replicate if all the attributes it is asking for
are in the GC partial attribute set
Autobuild-User: Andrew Tridgell <tridge@samba.org>
Autobuild-Date: Tue Sep 20 13:47:38 CEST 2011 on sn-devel-104
|
|
We previously only allowed objects of class ntDSDSA
Andrew Bartlett
Autobuild-User: Andrew Bartlett <abartlet@samba.org>
Autobuild-Date: Fri Aug 26 15:34:21 CEST 2011 on sn-devel-104
|
|
|
|
the NTDS entry
This entry has most of the time subelements (connections between DCs)
that will forbid a simple (non recursive) delete
|
|
loops when joining s4 domains
Autobuild-User: Matthieu Patou <mat@samba.org>
Autobuild-Date: Wed Jul 27 00:46:56 CEST 2011 on sn-devel-104
|
|
When an object is deleted, link pointed to it are marked as inactive.
When the same object is purged we do not remmove the link pointed to it
(we can't know them) so they stay in the database, it turns to be a
problem for Windows 2008.
Signed-off-by: Andrew Tridgell <tridge@samba.org>
|
|
this improved the logging of bugs related to linked attribute
replication
Autobuild-User: Andrew Tridgell <tridge@samba.org>
Autobuild-Date: Fri Jun 10 03:26:21 CEST 2011 on sn-devel-104
|
|
upgraded links can be missing the RMD_ADDTIME field
|
|
sub-function "forget" to do this
|
|
Right now it is solely based on function that handles objects
in normal DsGetNCChanges calls.
|
|
|
|
|
|
in case that destination_dsa_guid is not valid
|
|
|
|
"ldb_dn_validate" is NULL-safe and does the check implicitly.
Reviewed by: Tridge
|
|
Reviewed by: Jelmer
Autobuild-User: Matthias Dieter Wallnöfer <mdw@samba.org>
Autobuild-Date: Tue Mar 1 17:57:47 CET 2011 on sn-devel-104
|
|
The SPN attribute could derive from an untrusted source (client).
Reviewed-by: Jelmer
|