[Date Prev][Date Next]
Re: slapd connection delays
this might be a DNS issue. by default the server will try to do a reverse
lookup on the client connection coming in.
check your DNS setup and if you don't want to use DNS you could compile the
daemon yourself without the reverse lookup (see "./configure --help")
on 4/9/01 2:53, Noah Adams at firstname.lastname@example.org wrote:
> I am attempting to use ldap as an authentication backend for a network
> of linux machines. The machines are all running debian woody, and the
> ldap server is running the latest slapd in debian unstable. After
> configuring libnss-ldap and pam-ldap on the ldap server, everything
> seemed to be working fine; However, only on the server machine itself,
> all of the other machines encounter a timeout whenever they try to do
> anything involving ldap (chmodding a file to a user in the ldap tree,
> trying to login, etc.). Using netstat, i noticed that there was a
> connection during this period to the ldap server, but nothing seemed to
> happen on the client end until it hit the timeout, where it of course
> gives up.
> In fact, even running an ldap browser (gq) there is about a 30 second
> delay while connecting to the server, it seems to get stuck on a call to
> ldap_simple_bind_s(); for the duration of the delay. As I said before
> though, everything is nearly instantaneous on the server machine itself.
> The machines are all on a 10 Base-T lan, attached to the same hub in
> fact, and are all relatively new (nothing below 800 MHz) so I don't
> believe hardware is the problem, also, all other services (ssh, http and
> nfs) seem to work fine, so I don't believe it's the network either.
> Any help would be greatly appreciated,
> Noah Adams