Re: [Samba] ldbsearch/kerberos issue

2013-04-19 Thread Geoff Crompton

On 19/04/13 12:06, Geoff Crompton wrote:

Samba 4.0.5, Debian 6.0

I can successfully perform an ldbsearch on the Samba ldb by specifying
the -U parameter:
snip

and while I can kinit successfully, as this klist shows:
snip
I cannot use the resulting ticket to
connect:
snip
Anyone know what's going on? This email may look familiar, Steve
reported the same issue last July,
https://lists.samba.org/archive/samba/2012-July/168315.html. This isn't
the problem that Zach was talking about in
https://lists.samba.org/archive/samba/2012-November/169941.html, as I'm
not using an IP address in the url.



Someone contacted me offlist with a very helpful suggestion. Kerberos 
can't cope with a 'localhost' domain name.


So the following search works with kerberos authentication:

   $ /usr/local/samba/bin/ldbsearch -H
ldap://dc01.testad2.trinity.unimelb.edu.au -k yes 'CN=IT'

Cheers,
Geoff
--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


Re: [Samba] user both missing and present

2013-04-18 Thread Geoff Crompton

On 17/04/13 13:46, Geoff Crompton wrote:

I'm trying the classicupgrade, and it is crashing out:

snip

I've found that the user -42545 does not exist in my samba3 as a user,
but somehow does exist as a member of the -1231 group.

I'm using the ldapsam passdb backend, and I don't understand how this
situation has occurred. In my LDAP directory the -42545 user looks just
like some of my other users (who don't trigger this problem).

What should I do next?


I found that my ldap server was limiting the number of objects returned 
to searches. The script didn't get details for the -42545 user when 
asking about all users, which explains my problem.


 Once I set that to 'unlimited' I got past this problem. I've made a 
note on 
https://wiki.samba.org/index.php/Samba4/samba-tool/domain/classicupgrade/HOWTO#Notes_about_migrating_from_LDAP_backend 
for other people to watch out for that.


I wonder if the C library code behind the call to s3db.search_users(0) 
in the upgrade_from_samba3() method of python/samba/upgrade.py should 
handle an LDAP server telling it the error condition that it's search 
has been limited (and then the python adaption could throw that up as an 
exception).


I've created bug 9808 ( https://bugzilla.samba.org/show_bug.cgi?id=9808 
) along those lines.


Cheers,
Geoff
--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[Samba] ldbsearch/kerberos issue

2013-04-18 Thread Geoff Crompton

Samba 4.0.5, Debian 6.0

I can successfully perform an ldbsearch on the Samba ldb by specifying 
the -U parameter:



geoffc@test-dc03: ~ $ /usr/local/samba/bin/ldbsearch -H 
ldap://localhost  -U geoffc 'CN=IT' objectClass

Password for [STAFF\geoffc]:
# record 1
dn: CN=IT,CN=Users,DC=testad2,DC=trinity,DC=unimelb,DC=edu,DC=au
objectClass: top
objectClass: posixGroup
objectClass: group
snip

and while I can kinit successfully, as this klist shows:

geoffc@test-dc03: ~ $ klist
Ticket cache: FILE:/tmp/krb5cc_12823
Default principal: geo...@testad2.trinity.unimelb.edu.au

Valid starting ExpiresService principal
04/19/13 10:35:28  04/19/13 20:35:28 
krbtgt/testad2.trinity.unimelb.edu...@testad2.trinity.unimelb.edu.au

renew until 04/20/13 10:35:24
04/19/13 10:35:32  04/19/13 20:35:28 
ldap/dc01.testad2.trinity.unimelb.edu...@testad2.trinity.unimelb.edu.au



I cannot use the resulting ticket to
connect:

geoffc@test-dc03: ~ $ /usr/local/samba/bin/ldbsearch -H 
ldap://localhost -k yes  'CN=IT'
Failed to bind - LDAP client internal error: 
NT_STATUS_INVALID_PARAMETER

Failed to connect to 'ldap://localhost' with backend 'ldap': (null)
Failed to connect to ldap://localhost - (null)

Anyone know what's going on? This email may look familiar, Steve 
reported the same issue last July, 
https://lists.samba.org/archive/samba/2012-July/168315.html. This isn't 
the problem that Zach was talking about in 
https://lists.samba.org/archive/samba/2012-November/169941.html, as I'm 
not using an IP address in the url.


Cheers,
Geoff
--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[Samba] user both missing and present

2013-04-16 Thread Geoff Crompton

I'm trying the classicupgrade, and it is crashing out:

ERROR(class 'samba.provision.ProvisioningError'): uncaught exception -
ProvisioningError: Could not add member 'S-1-5-21-305843216-174882882-
1673945419-42545' to group 'S-1-5-21-305843216-174882882-1673945419-
1231' as either group or user record doesn't exist: Unable to find GUID
for DN
  File /usr/local/samba/lib/python2.6/site-
packages/samba/netcmd/__init__.py, line 175, in _run
return self.run(*args, **kwargs)
  File /usr/local/samba/lib/python2.6/site-
packages/samba/netcmd/domain.py, line 1318, in run
useeadb=eadb, dns_backend=dns_backend, use_ntvfs=use_ntvfs)
  File /usr/local/samba/lib/python2.6/site-packages/samba/upgrade.py,
line 919, in upgrade_from_samba3
add_users_to_group(result.samdb, g, groupmembers[str(g.sid)],
logger)
  File /usr/local/samba/lib/python2.6/site-packages/samba/upgrade.py,
line 316, in add_users_to_group
raise ProvisioningError(Could not add member '%s' to group '%s' as
either group or user record doesn't exist: %s % (member_sid, group.sid,
emsg))


I've found that the user -42545 does not exist in my samba3 as a user, 
but somehow does exist as a member of the -1231 group.


I'm using the ldapsam passdb backend, and I don't understand how this 
situation has occurred. In my LDAP directory the -42545 user looks just 
like some of my other users (who don't trigger this problem).


What should I do next?

Cheers,
Geoff
--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[Samba] python scripting samba

2013-04-15 Thread Geoff Crompton
Can someone point me to some documentation on scripting samba user and 
group management from python? I'd much rather not do this via calls out 
to samba-tool, and if I could do this remotely (via LDAP like calls) I'd 
be even happier.


Cheers,
Geoff
--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


Re: [Samba] tests failing on Debian Squeeze

2013-04-14 Thread Geoff Crompton
On 14/04/13 06:48, Matthieu Patou wrote:
 On 04/12/2013 05:11 PM, Geoff Crompton wrote:
 FAILED (318 failures, 7 errors and 0 unexpected successes in 431
 testsuites)

 A summary with detailed information can be found in:
./st/summary
 ERROR: test failed with exit code 1
 make: *** [test] Error 1



 How important is it for all the tests to pass? Should I just press on
 with a 'sudo make install'?

 No it's ok but in the same time if you could file a bug for each failing
 test so that we can start to have a look at fixing them.
 

Hi Matthieu,

Thanks for your response. The summary suggests there are over 300 failed
tests. Do you really want over 300 bug submissions? If so, I'll have to
work out how to automate that (maybe with
https://github.com/LegNeato/bztools).

What kind of extra information is useful along with these bug reports?
I'm assuming on each one you'd want the relevant section of the
./st/summary file, and mention the environment it was run in
(distribution, installed packages and their versions). Is there a script
in the samba tree to extract from the system this summary?

Cheers,
Geoff

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


Re: [Samba] tests failing on Debian Squeeze

2013-04-14 Thread Geoff Crompton

On 15/04/13 04:59, Andrew Bartlett wrote:

How important is it for all the tests to pass? Should I just press on
with a 'sudo make install'?


It's not good, but without seeing them it's rather hard to say more than
that.  Our tests are highly inter-dependent, so this is probably one or
two tests that really go awry, and 300 tests that fail because the
server process they run against has died.

Andrew Bartlett



I've submitted bug 9793, https://bugzilla.samba.org/show_bug.cgi?id=9793 
so you can have a look at the test results.


Cheers,
Geoff
--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[Samba] file share necessary?

2013-04-14 Thread Geoff Crompton
On 
https://wiki.samba.org/index.php/Samba_AD_DC_HOWTO#Setup_a_basic_File_Share 
the instructions says


 For the server to be useful you, will need to update it to have at 
least one share



What do you need a file share for the server to be useful? Isn't a 
domain controller 'useful' to authenticate machines and users even if it 
doesn't have a file share? I would have thought that this step would be 
optional, like the NTP step just above it.


Cheers,
Geoff
--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


Re: [Samba] file share necessary?

2013-04-14 Thread Geoff Crompton

On 15/04/13 14:07, Marc Muehlfeld wrote:

Am 15.04.2013 04:23, schrieb Geoff Crompton:

On
https://wiki.samba.org/index.php/Samba_AD_DC_HOWTO#Setup_a_basic_File_Share

the instructions says

For the server to be useful you, will need to update it to have at
least one share


What do you need a file share for the server to be useful? Isn't a
domain controller 'useful' to authenticate machines and users even if it
doesn't have a file share? I would have thought that this step would be
optional, like the NTP step just above it.


The DC need at least a sysvol and netlogon share.


Regards,
Marc


That makes sense. If you've followed the HOWTO (as I just have) you 
probably already have sysvol and netlogon shares. Does anyone mind if I 
mark the 'Setup a basic File Share' as optional, and add some notes that 
the sysvol and netlogon shares are needed for a domain controller, but 
users should feel free to add their own if they desire?


Cheers,
Geoff
--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[Samba] tests failing on Debian Squeeze

2013-04-12 Thread Geoff Crompton
Hi,

I'm trying to get samba4 4.0.5 tarball to run on a Debian Squeeze
Virtual Machine. I've done a local build, and I've included the
--enable-debug --enable-selftest switches.

I've done a 'make test', and I get told there are quite a lot of
failures. It ends by saying:


FAILED (318 failures, 7 errors and 0 unexpected successes in 431
testsuites)

A summary with detailed information can be found in:
  ./st/summary
ERROR: test failed with exit code 1
make: *** [test] Error 1



How important is it for all the tests to pass? Should I just press on
with a 'sudo make install'?

Cheers,
Geoff
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba


[Samba] whats normal for samba loging amount

2005-03-30 Thread Geoff Crompton
Hi,
I'm running the Debian samba 3.0.10-1 package. I have 'log level = 0'. 
However my logcheck is showing a lot of noise from samba. Are the 
following messages normal, or do we have a misconfiguration? (We don't 
seem to have any problems accessing the shared samba shares):

Mar 31 12:13:31 sd01 nmbd[3150]: [2005/03/31 12:13:31, 0] 
nmbd/nmbd_browsesync.c:get_domain_master_name_node_status_fail(488)
Mar 31 12:13:31 sd01 nmbd[3150]:   get_domain_master_name_node_status_fail:
Mar 31 12:13:31 sd01 nmbd[3150]:   Cannot get workgroup name.
Mar 31 12:25:36 sd01 smbd[14299]: [2005/03/31 12:25:36, 0] 
lib/util_sock.c:get_peer_addr(1000)
Mar 31 12:25:36 sd01 smbd[14299]: [2005/03/31 12:25:36, 0] 
lib/util_sock.c:get_peer_addr(1000)
Mar 31 12:25:36 sd01 smbd[14299]: [2005/03/31 12:25:36, 0] 
lib/util_sock.c:write_socket_data(430)
Mar 31 12:25:36 sd01 smbd[14299]: [2005/03/31 12:25:36, 0] 
lib/util_sock.c:write_socket(455)
Mar 31 12:25:36 sd01 smbd[14299]:   write_socket: Error writing 4 bytes 
to socket 24: ERRNO = Connection reset by peer
Mar 31 12:25:36 sd01 smbd[14299]: [2005/03/31 12:25:36, 0] 
lib/util_sock.c:send_smb(647)
Mar 31 12:25:36 sd01 smbd[14299]:   Error writing 4 bytes to client. -1. 
(Connection reset by peer)
Mar 31 12:28:32 sd01 nmbd[3150]: [2005/03/31 12:28:32, 0] 
nmbd/nmbd_browsesync.c:get_domain_master_name_node_status_fail(488)
Mar 31 12:28:32 sd01 nmbd[3150]:   get_domain_master_name_node_status_fail:
Mar 31 12:28:32 sd01 nmbd[3150]:   Cannot get workgroup name.
Mar 31 12:48:32 sd01 smbd[12738]: [2005/03/31 12:48:32, 0] 
smbd/connection.c:register_message_flags(220)
Mar 31 12:48:34 sd01 smbd[12738]: [2005/03/31 12:48:34, 0] 
smbd/connection.c:register_message_flags(220)

Thanks for any comments
--
Geoff Crompton
Debian System Administrator
Strategic Data
+61 3 9340 9000
--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba