Testing Samba4 vampire

[Stefan (metze) Metzmacher]

Andrew Bartlett schrieb:
> Currently, our Samba4 'vampire' and DRS replication code is not well
> tested.  We do conduct regular manual testing, but frankly nothing beats
> an automated test - even of Samba4 to Samba4.
> 
> In the 'vampire-test-wip' branch in my git repository (and in my recent
> commits to master), you will find the work I've done to try and make
> this work.
> 
> git://git.samba.org/abartlet/samba.git vampire-test-wip
> 
> There are many bugs and quandaries - why do NET-API-* tests fail when we
> have the vampired DC, but not otherwise?

The last time I've tried your branch, the replication was failing
(the 2nd dc wasn't able to get a rid set). And the NET-API-* tests
connect to any dc and fail to create users against the 2nd dc.

> Why do we, despite all the work
> to update the schema, still get schema errors in the DRS server?
> 
> If you have the time or inclination, please do try a 'make test' in that
> branch, and see if you can help find these bugs. 

We don't correctly handle the case where we need to first replicate the
schema partition before we can replicate the other partitions.

metze


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 260 bytes
Desc: OpenPGP digital signature
URL: <http://lists.samba.org/pipermail/samba-technical/attachments/20100322/730a563d/attachment.pgp>