Re: [ldapext] DBIS - new IETF drafts

[Michael Ströder <[e-mail]>]

Simo wrote:
> On Thu, 2014-01-09 at 16:48 +0100, Michael Ströder wrote:
>> Mark R Bannister wrote:
>>> Yes as you'll see from my recent reply to Simo, the case sensitivity issue was
>>> one of the problems I faced at a large installation,
>>
>> This could be easily solved with RFC2307bis. Or not?
>>
>> In my deployments I simply define additional (OpenLDAP) constraints for those
>> attributes (e.g. to enforce lower-case 'uid' values).
>>
>> IMHO only re-defining the matching rules does not fully solve the case problem
>> anyway. Restricting to lower-case attribute values helps better.
> 
> 
> I'd go beyond this, supporting case-sensitive user names is actively
> harmful for various reasons.
> 
> - Assuming users (and admins) should be able to distinguish based on
> case is wrong, we naturally consider the strings 'Admin' and 'admin' to
> be the same thing.
> 
> - Some systems are case-preserving (meaning they'll show you back the
> same case you entered, but are really case-insensitive and if you have
> to interoperate with them you cannot assume Admin and amdin to be
> different users, it could lead to serious security issues.
> 
> - If we are in the legacy game, there are still systems that will simply
> accept only all caps names, like ADMIN. In these cases what do you map
> that to ? Admin ? admin? a third user called ADMIN ?
> 
> And there are many other examples where really being case sensitive
> causes a lot more problem than it resolves.
> Due to these problems what we did in FreeIPA is to always create users
> in lower case and explicitly state we are case-preserving and
> insensitive. It is the only reasonable compromise IMHO.

So basically you're doing the same thing like I described above.

>>> Nested groups are very important especially in large enterprises, and really
>>> do assist with data management.
>>
>> I am always getting told this but I have strong doubts about nested groups.
> 
> Great value at least in IPA, we use them a lot to associate things like
> permissions to roles to groups and ultimately to users (All through
> nesting each of these as groupsofnames). It does really make some
> problems much easier to handle at the end of the process.

Which problems are easier? Personally I'd definitely consider abusing
groupOfNames for permissions bad schema design.

>>>  Yes we'll get more search operations, but I
>>> don't think this will be a problem.  I should be able to prove this will work
>>> in my reference implementation.
>>
>> Resolving nested group membership is a big performance cost. I can see this
>> with a MS Sharepoint installation working with a OpenLDAP server. Sharepoint
>> sends many search requests even though nested groups are not used in this
>> deployment.
> 
> Performance issues with nested groups can easily be solved via caching
> and the deref control though.

But the deref control gives you only one level. Speaking of nested groups in
general people mean more than just two-level group memberships.

Ciao, Michael.