[Date Prev][Date Next]
Re: DSE Root/first entry
... the thread moved from openldap-software to openldap-devel...
"Kurt D. Zeilenga" wrote:
> At 10:29 AM 10/13/00 +0100, Konstantin Chuguev wrote:
> >You say that with OpenLDAP we should be able to manage suffix "" and first level entries, only
> >a bug prevents it. I would like to help to find the bug. Can I know the details please.
> Basically, one needs to special case a base DN "" search
> such that it the search "locates" the base. Then likely
> the DN index subsystem would need to be adjusted.
> There are some changes in HEAD which might help... they
> need to be tested and any other issues raised, discussed,
> and resolved.
Sorry for being too curious :-)
Can anybody give me more details on what exactly is the problem? Looking through the sources I
could see that rootDN-awareness is already implemented. Before I go to more detailed
debugging/patching/playing, could developers confirm I am on a right way.
For each entry in LDBM, there are three index entries: "=<DN>" for the entry itself (for base
search); "%<DN>" for the entry's immediate parent (for one-level search) and "@<DN>" for all
entry's parents up the tree. Last two cases allow existence of "%" and "@" index keys. The former
points to all the first level entries' LDBM IDs, the latter to all the IDs in the LDBM database.
It looks like dn2id.c:dn2idl(), called from filterindex.c:filter_candidates(), called from
search.c:search_candidates(), called from search.c:ldbm_back_search(), is pretty OK.
Can it be that this code (i was looking at 2.0.4) is those changes in HEAD Kurt was talking about?
Anyway, I could not find any recent changes regarding root DN on CVSweb:
So, if this is the code responsible for root DN processing, I am going to configure and test my
server. But if I am totally wrong, I would really appreciate some hints.
* * Konstantin Chuguev - Application Engineer
* * Francis House, 112 Hills Road
* Cambridge CB2 1PQ, United Kingdom
D A N T E WWW: http://www.dante.net