[PATCHES] idmap_rfc2307 with test

Andrew Bartlett abartlet at samba.org
Tue Mar 5 15:10:05 MST 2013


On Tue, 2013-03-05 at 19:39 +1100, Andrew Bartlett wrote:
> On Mon, 2013-03-04 at 16:04 -0700, Christof Schmitt wrote:
> > On Sat, Mar 02, 2013 at 11:35:04AM +1100, Andrew Bartlett wrote:
> > > On Sat, 2013-03-02 at 10:29 +1100, Andrew Bartlett wrote:
> > > > Some more challenges:
> > > > 
> > > > called with: SAMBADOMAIN Administrator 2000000 Domain Users 2000001
> > > > ou=idmap,dc=samba,dc=example,dc=com
> > > > UNEXPECTED(failure): idmap.rfc2307.add ldap prefix(s3member_rfc2307)
> > > > REASON: _StringException: _StringException: ltdb: tdb(./st/dc/private/sam.ldb): tdb_open_ex: could not open file ./st/dc/private/sam.ldb: No such file or directory
> > > > 
> > > > Unable to open tdb './st/dc/private/sam.ldb'
> > > > Failed to connect to './st/dc/private/sam.ldb' with backend 'tdb': Unable to open tdb './st/dc/private/sam.ldb'
> > > > Failed to connect to ./st/dc/private/sam.ldb - Unable to open tdb './st/dc/private/sam.ldb'
> > > > 
> > > > The tests fail autobuild (you can run an autobuild locally
> > > > with ./script/autobuild.py), because you assume ./st. 
> > > > 
> > > > 
> > > > We want the ability to change that in the future, so you need to use
> > > > things like the PREFIX variable to find the files, or connect to the DB
> > > > over LDAP (this we prefer because it makes it easier to run the same
> > > > test against a real windows DC).  (The path in autobuild is ./bin/ab to
> > > > trip up this kind of mistake).
> > > > 
> > > > Can you fix that up for me, and re-submit?
> > > 
> > > Also please delete the objects added once the test is over, and delete
> > > them before the test starts.  This will allow the test to be run
> > > multiple times in 'make testenv' without difficulty, when debugging
> > > issues.
> > 
> > Thanks for pointing out those issues. Here is the updated patch
> > series. I removed the LD_LIBRARY_PATH patch, since it is not needed.
> > The testcase now connects to the LDAP server instead of accessing the
> > local file, and it deletes the records before and after the test.
> > 
> > Regards,
> 
> Did this pass a local autobuild for you?
> 
> I tried to get it into master, and the autobuild on sn-devel failed
> with:
> 
> Connection to \\LOCALADMEMBER\tmp failed - NT_STATUS_LOGON_FAILURE
> UNEXPECTED(failure): samba4.blackbox.smbclient(s3member:local).Test login with --machine-pass without kerberos(s3member:local)
> REASON: _StringException: _StringException: No reason specified
> 
> FAILED (1 failures, 0 errors and 0 unexpected successes in 0 testsuites)
> 
> A summary with detailed information can be found in:
>   ./bin/ab/summary
> ERROR: test failed with exit code 1
> 
> 
> (Please don't feel bad about it, I've fed Jeremy a bung patch series
> trying to add a simple command line option twice so far, so it happens
> to all of us).

I can't fully pin this down, but I'm suspecting an interaction between
tests, perhaps because this essentially adds a duplicate user object
(with the idmap stuff on it).  

Either way, can we create two users and groups, one with just put the
idmap info on the main user/group object, and other for the separation?
Then just have those users alive for only the length of the idmap
test?  

Otherwise, I just wish you the best of luck working out the collision
between the test (a similar issue took me two hours last night, they can
be very complex to resolve). 

Thanks,

Andrew Bartlett

-- 
Andrew Bartlett                                http://samba.org/~abartlet/
Authentication Developer, Samba Team           http://samba.org




More information about the samba-technical mailing list