[SCM] Samba Shared Repository - branch master
steven.danneman at isilon.com
Fri Feb 13 12:16:18 MST 2009
> -----Original Message-----
> From: Stefan (metze) Metzmacher [mailto:metze at samba.org]
> Sent: Thursday, February 12, 2009 11:39 PM
> To: Steven Danneman
> Steven Danneman schrieb:
> >> On Wed, Feb 11, 2009 at 09:39:30PM -0600, Steven Danneman wrote:
> >>> commit d8c54fddda2dba3cbc5fc13e93431b152813892e
> >>> Author: Dan Sledz <dsledz at isilon.com>
> >>> Date: Tue Feb 10 15:50:39 2009 -0800
> >>> s3: Change behavior when seeing an unknown domain.
> >>> After a lot of testing against various Windows servers (W2K,
> >> W2K3, W2K8),
> >>> within an AD domain it seems that unknown domains will only be
> >> translated
> >>> to the local account domain, not the netbios name of the
> >> server's
> >>> domain. This makes samba act more like Windows.
> >> This one has caused a great deal of discussion in the past.
> >> I'm with you on the semantics, but at the point where this was
> >> discussed I eventually gave up pushing it through. I don't want to
> > stir
> >> up old stories again, but all who had stakes in this might want to
> >> speak up again.
> >> And -- can you please next time make one reformatting patch and one
> >> semantics changing patch out of this? Or better:
> >> Don't do reformatting at all...
> > Hey Volker,
> > Agreed, this patch is a modified (and more educated) version of the
> > patch I submitted to do the same thing back in Sep 2007
> > (http://www.nabble.com/Mapping-workstation%5Cuser-to-domain%5Cuser-
> > or
> > rectly--td12808573.html#a12808573) and I know there was more
> > discussion before that.
> > I can say that Isilon has been running with this patch for the last
> > 1.5 years on all customer sites AND this is the demonstrable
> > of Windows in all of our testing.
> > I concede that some people may be relying on this behavior and would
> > be willing to parameterize it with the default to Windows behavior
> > there's strong dissension?
> > On the cleanup, yes you're right I'll separate them into different
> > patches next time.
> Hi Tim,
> It would be nice to make some torture tests for this behavior, (or at
> least blackbox tests running smbclient with different -U options in
> make test)
I didn't do this initially because I wasn't sure what infrastructure was
in place in make test to first setup a DC and setup Samba as member
I remember there was some discussion of doing this automatically using
Samba4 as the DC. Could you point me to any tests that have similar
prerequisites that I could model off of?
More information about the samba-technical