In case someone was using smbldap-tool 0.85, this change is around line 390, and looks like below (note there is no "$ldap_master->unbind;" and there is "return 1;") - I set it to 15, just in case:
Hrm, coulda sworn that I was using 0.85... but I have been wrong before, just once.
It should be noted that you should make sure that LDAP is your issue before doing this just so that you aren't masking the real issue. In my case you could watch the LDAP query come through while watching the samba logs and you'd actually see samba asking for the sambaSAMAccount entry before it replicated through to the slave. This sounds like your issue as well if your 'high latency link' is slow enough.
If you're running a simple non-replicated setup or if your slave isn't over a high(ish) latency link I'd keep looking for other issues.
I spent all yesterday thinking what can be wrong, and today the whole day trying to figure out by changing different settings / watching logs etc.
I googled for people with similar problem, but there weren't many, the only solution to the problem I had was this "sleep" added to smbldap-tools (thanks for that).
In logs I could see that Samba is complaining that it can't find the machine/name [it just added] (writing happens to a remote master over worst case slow ADSL/VPN link, then it's replicated to a slave over the same link), so I think that approach with adding "sleep" is good.
But finally it works; sometimes I was thinking that these M$ guys talking about higher Linux TCO might be right :)
Tomek
-- To unsubscribe from this list go to the following URL and read the instructions: http://lists.samba.org/mailman/listinfo/samba
