3 Replies Latest reply: Oct 5, 2012 1:36 PM by atomicboy
atomicboy Level 1 Level 1 (5 points)

I have been having a problem with one of the Lion servers I support which is I cannot use either of the GUI tools to administer it. This does not appear to be the problem others have had with Lion where the servermgrd process isn't running. For example:

 

acropolis:~ ladmin$ sudo launchctl load -w /System/Library/LaunchDaemons/com.apple.servermgrd.plist

Password:

bind(): Address already in use

bind(): Address already in use

com.apple.servermgrd: Already loaded

 

Instead this seems to be that the login information I supply to either the Server app or the Server Admin app is being applied against the wrong directory so the user isn't being authenticated so the apps cannot connect. It is also possible I simply have this server mis-configured but now I am stuck. Unless there is something I can do from the command line to fix I can't fix it. This is a mission critical server so any restarts etc. have had to wait for scheduled maintenance windows.

 

This machine is an Xserve running Lion 10.7.5. It hosts DHCP, DNS, AFP shares, as well as being an OD Master. However, it is importing users and groups from our old OD Master (running 10.6.8) for the time being. This arrangement worked OK for several months. I have been looking for a large window where everything could basically be disrupted while the old Snow Leopard server is decommissioned and the user accounts are set up on the Lion server. Kerberos is suspect here too since the realm is still tied to the old server (but I haven't seen any errors related to kerberos). Below are some of the clues I think I've found.

 

It seems that when I try and connect either of the Server apps it tries the "local LDAP node" which I thought would be /Local/Default but instead I think its trying to authenticate against /LDAPv3/127.0.0.1 which doesn't have any of the local users.

 

10/2/12 1:58:09.493 PM servermgrd: servermgr_accounts: got error 5000 trying to auth to local LDAP node

 

I see errors related to the GlobalGID as well as the GroupName. Is this because of the groups being imported from the old OD Master?

 

10/2/12 1:02:51.209 PM opendirectoryd: Misconfiguration detected in hash 'GlobalGID' - see /var/log/opendirectoryd.log for details

10/2/12 1:02:51.234 PM opendirectoryd: Misconfiguration detected in hash 'GlobalGID' - see /var/log/opendirectoryd.log for details

 

10/2/12 10:20:51.710 AM opendirectoryd: Misconfiguration detected in hash 'GroupName' - see /var/log/opendirectoryd.log for details

10/2/12 10:20:51.728 AM opendirectoryd: Misconfiguration detected in hash 'GroupName' - see /var/log/opendirectoryd.log for details

 

When I look through /var/log/opendirectoryd.log I find entries like this that repeat over and over:

 

2012-09-28 23:25:37.189 PDT - Module: SystemCache - Misconfiguration detected in hash 'GroupName':

 

So I am really stuck and any clues or guidance would be really helpful.

 

Message was edited by: atomicboy