linking 2 hub error

[Jahangir]

First hub running latest versions on UnrealIRCd, Anope and BOPM.

Code: Select all

	name "irc1.asian-flava.com";
	info "Asian Flava";
	numeric 1

Code: Select all

listen 72.20.23.93:6667 {
	options {
		java;
		clientsonly;
	};
};
listen 72.20.23.93:7029 {
	options {
		serversonly;
	};
};
listen 72.20.23.93:7030 {
	options {
		serversonly;
	};
};

Code: Select all

link            services.asian-flava.com
{
	username	*;
	hostname 	72.20.23.93;
	bind-ip 	*;
	port 		7029;
	hub             *;
	password-connect "password";
	password-receive "password";
	class           servers;
};

link irc2.asian-flava.com {
	username *;
	hostname 188.165.75.168;
	bind-ip *;
	port 7030;
	hub *;
	password-connect "password";
	password-receive "password";
	class servers;
	options {
		autoconnect;
		nodnscache;
		nohostcheck;
	};
};

Second hub running latest versions on UnrealIRCD and BOPM.

Code: Select all

	name "irc2.asian-flava.com";
	info "Asian Flava";
	numeric 3;

Code: Select all

listen 188.165.75.168:6667 {
	options {
		java;
		clientsonly;
	};
};
listen 188.165.75.168:7029 {
	options {
		serversonly;
	};
};
listen 188.165.75.168:7030 {
	options {
		serversonly;
	};
};

Code: Select all

link            services.asian-flava.com
{
	username	*;
	hostname 	72.20.23.93;
	bind-ip 	*;
	port 		7029;
	hub             *;
	password-connect "password";
	password-receive "password";
	class           servers;
};

link irc1.asian-flava.com {
	username *;
	hostname 72.20.23.93;
	bind-ip *;
	port 7030;
	hub *;
	password-connect "password";
	password-receive "password";
	class servers;
	options {
		autoconnect;
		nodnscache;
		nohostcheck;
	};
};

have no problems when i connect to irc2.asian-flava.com but when i connect to irc1.asian-flava.com i get this error every 30secs

-irc1.asian-flava.com- *** Notice -- Link [@89.238.135.213.61575] cancelled, server irc2.asian-flava.com already exists from irc2.asian-flava.com

i tried without the autoconnect option on the first hub but makes no difference

[katsklaw]

the IP in the error doesn't match either irc1 or irc2, that means there is another ircd somewhere trying to connect as irc2

[Jahangir]

irc2 was allready running on a different shell its sorted now now i want to link 2 more leafs to it, might add more in the future. how do i set it up so irc3 (leaf) to connect to irc1 but when irc1 is down it connects to irc2 and irc4 (leaf) to connect to irc2 but when irc2 is down it connects to irc1?

[katsklaw]

irc2 was allready running on a different shell its sorted now now i want to link 2 more leafs to it, might add more in the future. how do i set it up so irc3 (leaf) to connect to irc1 but when irc1 is down it connects to irc2 and irc4 (leaf) to connect to irc2 but when irc2 is down it connects to irc1?

Nothing wrong with using leafs and hubs, to do this read: http://www.unrealircd.com/files/docs/un ... #linkblock

However, I recommend that all servers be compiled and configured so they can be linked anywhere. It can definitely save a headache later. This way if irc3 is gone irc4 isn't in purgatory until someone with shell access can add a link block. I recommend all servers having link blocks for all other servers. This way if irc3 drops, you simply type: "/connect irc4.* <port> <server irc4 is to link to>" and no headache about where and when irc4 can return to the tree.

Second;y, there is no automated primary/secondary linking schema. You can have a server set to autoconnect in the link block, but that's about the limit of things. This is where your IRCOps will actually have to do the job they are there to do. The IRCops job is not babysitter for users and channels, contrary to today's practice. The job the IRCop or IRC Operator is to *gasp* Operate the IRC Server. This means IRCops are expected to know how to use the /connect command and others that relate to routing and re-routing network servers. The /kill and /kline commands were added to help deal with unruly users, not as a precursor to customer service.

Hope that helps.