This content has been marked as final. Show 1 reply
This is a behaviour which follows the same pattern as the predecessors NIS and NIS+.
I think that its an heritage from the time when Solaris was a client OS; on a client this sort of behaviour makes sort of sence, since noone will be able to use a client or workstation if NIS/NIS+/LDAP is down. On a server however its quite disastrous.
Whilst at Sun i filled a bug on this behaviour; 5094148. I think it was fixed in Solaris 10 as a part of the SMF implementation.
I dont think there is any good way of solving this problem in Solaris 9, but i think you can solve this problem by going to the /etc/init.d directory, grep for ldap_cachemgr and then edit the scripts that starts it and adding an ampersand (&) after 'ldap_cachemgr'. I don't have any Solaris 9 box to check the exact file though.