[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: ACL filter not working

Thanks! That's exactly what I was missing, looks like some default changed between OpenLDAP 2.3 and 2.4. I added the correct search permissions as you suggested and that got it working.


On 01/11/2010 2:02 PM, Jonathan Clarke wrote:
Le 01/11/2010 17:56, Darren Hildebrand a écrit :
I'm having a problem with access control lists in slapd.conf. The filter
doesn't seem to be working in OpenLDPA 2.4.23 using syntax that worked
in 2.3.43. I've simplified my tests down to a single ACL rule just to
see if it's working, and this is what I'm finding:

I'm trying this ACL:

access to dn.subtree="ou=users,dc=companyname,dc=com"
by * read

I tried an anonymous search using this command:

ldapsearch -h -x -b ou=users,dc=companyname,dc=com

And it returned the following:

# search result
search: 2
result: 32 No such object

I tried the same with the same ACL as above with the filter set to
"(uid=*)", and got the same problem (note that all users have a uid
value set). However, I tried with the filter set to "(objectClass=*),
and it returned all users as I would expect.

Has anything changed with ACL syntax between OpenLDAP 2.3 and 2.4? Or do
you see any problems with the syntax of my ACL line above?

Yes, from slapd.access(5):
The search operation, requires search (=s) privileges on the entry pseudo-attribute of the searchBase (NOTE: this was introduced with OpenLDAP 2.4). Then, for each entry, it requires search (=s) privileges on the attributes that are defined in the filter. The resulting entries are finally tested for read (=r) privileges on the pseudo-attribute entry (for read access to the entry itself) and for read (=r) access on each value of each attribute that is

So you likely need something like this too:
access to dn.base="ou=users,dc=companyname,dc=com" attrs="entry"
        by * search

However, I'm not entirely sure why you're allowed to do the search when using the "(objectClass=*)" filter, even though you don't have access to the above-mentioned entry pseudo-attribute...