2) When for test I shutted nodo1 query and authentication was made >
correctly from nodo2, but I saw that in this case it was slower than with
> nodo1...for
every query I ave to wait about 3 seconds....while with nodo1 it was
istantaneous.
This is currently how PADL nss_ldap works. Whenever it needs to talk to a
ldap server, it tries the server list in order. It doesn't maintain state
about which servers in the list are down. If the first server in the list
is up, you'll get the fastest response.
The bind timelimit specifies how long it will try binding to each server in
the list. Setting this lower will give you better response time if the
first server is down.
There are some comments in the notes for PADL's latest nss_ldap release
about "more robust failover support". If you like, you could try the latest
nss_ldap version instead of the RedHat/Fedora bundled version and see what's
changed.
LDAP client support for server lists/failover varies a lot. Several client
apps don't implement it at all. So depending on what apps you need to
support, it often makes more sense to provide redundancy at the server
(e.g., with a load balancer).
----- Original Message -----
From: "Alex" <magobin@xxxxxxxxx>
To: "'General discussion list for the Fedora Directory server project.'"
<fedora-directory-users@xxxxxxxxxx>
Sent: Saturday, April 15, 2006 2:58 AM
Subject: NSS_ldap....(was... SubjectAltName how
does it work?)
Hi, following Susan's council...I leave out my purpose about floating ip
and
I set up two server ldap in client's configuration...int this way
everything
works, but I have two question:
1) After set up via authconfig in a fedora 5 client ldap configuration I
reboot client but on booting seems that something doesn't work...I have to
wait a lot of time..after this time X doens't work and I remake
Xconfiguration..but, after that I succesfully log in in client and query
on
ldap servers...I found in logs this:
Xfs: nfs_ldap: reconnetting to ldap server (sleeping 4 seconds)...
-CUT-
Xfs: nfs_ldap: reconnecting to ldap server (sleeping 40 seconds)...
And so on...but repeat...after X reconfiguration everything works..the
problem is every time I reboot client I have to configure X :-(
2) When for test I shutted nodo1 query and authentication was made
correctly
from nodo2, but I saw that in this case it was slower than with
nodo1...for
every query I ave to wait about 3 seconds....while with nodo1 it was
istantaneous.
Thanks for help
Alex
--
Fedora-directory-users mailing list
Fedora-directory-users@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/fedora-directory-users
--
Fedora-directory-users mailing list
Fedora-directory-users@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/fedora-directory-users
