[Date Prev][Date Next]
Re: Subtree delete not supported: HDB bug?
Eric Irrgang wrote:
The value of the hasSubordinate may be a symptom of the problem, but I
don't think it's the cause, because its value is generated by asking the
underlying BerkeleyDB dn2id database if the entry may have children.
Your second interpretation of the X.500 wording is consistent with ours,
i.e. hasSubordinate simply gives an indication whether an entry may or
may not have subordinates at the time the request was generated. That
is, if you search an entry with hasSubordinates=FALSE with oneLevel
scope, you may still find some subordinates and vice-versa.
What is the value of the hasSubordinates attribute is for some entries
that you can't delete? For some reason, most (but not all!) of my
ou=people entries report hasSubordinates=TRUE, though of course none of
them actually do have subordinates.
I notice the wording of one of the x500 specs as saying that
hasSubordinates=TRUE means that an entry "may" have subordinates. I don't
know whether that means the entry is allowed to have subordinates (but
doesn't LDAP allow all entries to have subordinates?) or whether it is
possible that the entry has subordinates and so further checks should be
performed to be sure? I don't know enough about how LDAP schema work to
be able to contribute meaningfully in this area...
SysNet - via Dossi,8 27100 Pavia Tel: +390382573859 Fax: +390382476497