Error 8418: The replication operation failed because of a schema mismatch between the servers involved

[Andrew Bartlett]

On Thu, 2016-03-31 at 21:46 +0300, Evgeny Sinelnikov wrote:
> 
> Ok, I re-work patch with dsdb_attribute_get_attid() without check
> that
> is schema attribute. This is looks like enough. If it is not I have
> another tested patch adds is_schema_nc argument to
> get_nc_changes_add_la() and get_nc_changes_add_links() functions.

We need that additional argument to be provided, and correctly filled
in.

> At finally, we need a manual test with replication for normal
> attributes from custom schema. We looks to ndrdump packets with it
> and
> inspect it.
> This is not torture reproducable test. Is it right?

I think we can do this with a automated test, and that we should.  The
same applies to other aspects of our schema replication, because we
have had so many issues here over time.

We must move to a situation where we track the attributeID values in
replPropertyMetaData, in the replicated object and in linked attributes
and assert that they are correct via an automated script.  The lack of
that testing is what has caused so much pain here, since long before
Samba 4.0 was released.

(This is why Samba refuses to accept schema modifications by default,
because we know this stuff isn't great).

I'm going to look into parts of this to fix some other issues, and that
will hopefully give you a framework for this final item.

Thanks,

Andrew Bartlett

-- 
Andrew Bartlett
https://samba.org/~abartlet/
Authentication Developer, Samba Team         https://samba.org
Samba Development and Support, Catalyst IT   
https://catalyst.net.nz/services/samba