Good morning!
In the interests of speeding up our moodle, today's topic: LDAP!
The context that we're searching is, well, everything in the tree. We're going to narrow it down, and we can also tune the LDAP server itself, but I'm curious about the algorithm. I can narrow it down to all the users in the tree, but I could also narrow it down to a per-group level, and then put each of those groups in the order that it is most likely to find the user.
But that only makes sense if it queries one context, fails, and rolls over to the next. It might fire off parallel queries instead. I did attempt to read the code, but given the whole abstraction process of M2.x it is far too complex for me without a tracer. And besides, one of you people already knows this.
And as long as I'm asking, does it just do a failover to the second or third server for queries, or does it do some sort of round-robin. I suppose it could even do round-robin and parallel queries.
Thanks.
by david millians.