On 7 Jul 1999 michael@??? wrote:
> > So: It appears that the facility is already present, at least at the
> > ldap_open() level, though strictly this breaks the rules for the URL.
>
> I know, but OpenLDAP views that as a bug that will be fixed in the
> future, because it violates RFCs.
Ah. It's a pity that the existing facility in ldap_open() cannot be
accessed in this simple way.
> My patch has been discussed on the list for a while and the result was
> that it may be wise to use a variable ldap_hosts, which holds a list of
> hosts to try, and unqualified LDAP URLs. That way we do not break the
> RFC and the URLs get shorter.
Yes, I saw all that, but wanted to be sure I understood the situation
properly before actually doing anything. The simple thing to do is
to just copy that list (ldap_hosts - actually I prefer
ldap_default_hosts) as the argument for ldap_open().
> My mail system depends on the ability to use redundant LDAP
> servers, because by now it is entirely controlled by LDAP (including
> local_domains). Example: I change the schema and re-initialise a
> local replica server. Copying the new data base to it takes a while.
> It would not be acceptable to defer mail during that time.
Does this mean that you want to be able to retry after bind failures and
timeouts to the reloading server? Or does it refuse connections when it
is reloading? If it refuses connections, then the simple approach of
handing the list to ldap_open() is all that is needed. If not, then
something like your patch has to be done, but I think it should really
be more complicated, because I don't think it should retry after an
error like "too many entries".
So what I'm asking still stands: do you need to retry for defers other
than those caused by ldap_open()?
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
--
*** Exim information can be found at
http://www.exim.org/ ***
