[Date Prev][Date Next]
slapd future questions
I've seen several threads on this list about whether to
use the LDBM or BDB backends for slapd. The last one
"LDBM verse BDM" had Frank Swasey saying LDBM wasn't
receiving much/any development. This brings up the
1. Is it true that LDBM isn't receiving development?
2. How long will LDBM be supported?
3. Is the BDB backend considered stable enough
for production environments?
4. Are there plans to move away from the BDB backend to
some new backend in the known future?
5. How long will the BDB backend be supported?
I'm asking these questions to find out what the developers
are planning for the future of OpenLDAP so I can deploy a
setup with some life ahead of it.
Also, I've tried compiling OpenLDAP threaded as well as
non-threaded. For the kind of queries I ran (indexed
searches) I found the non-threaded server to be faster,
even on a machine with 2 physical CPU's with hyperthreading
(so the OS used them as 4 CPU's). Which leads to more
6. Do people recommend running slapd without threads?
7. Will running slapd without threads be supported in the
future or is it a "your mileage may vary" thing?
In general I feel that if any of the aforementioned
options is not really supported (or even deprecated),
the documentation or configure should warn against
Wout van Albada