Good day. I need help very-very much with my Ejabberd cluster.
I had ejabberd-cluster of 2 machines, all was great. I decided to reinstall one of these machines, stopped it and tried to do everything again. I used the same conf-files, base, name, .erlang.cookie and even IP-address. But I can't get them in a cluster now! I just can't imagine how I did it previously. The last thing I could reach was this: I started 'ejabberdctl live' on both machines, I can see that cookies are the same, I can net_adm:ping one machine from another. But: on the 'First' machine I have:
- net_adm:world() - shows 'Second'
- nodes() - shows 'Second'
- running db nodes - shows only 'First'
- stopped db nodes - is empty
on the 'Second' machine I have:
- net_adm:world() - is empty
- nodes() - shows 'First'
- running db nodes - shows only 'Second'
- stopped db nodes - shows 'First' machine
I just can't understand what's the problem. Both machines can net_adm:ping each other, therefore cookies are the same (get_cookie also shows that they are the same). Why Second machine see First in nodes() but can't see in net_adm:world()? why First doesn't list Second as at least 'stopped db node'? I just broke my mind( Please help.. I used these tutorials:
I've got Debian Squeeze, ejabberd-2.1.9 compiled from sources.