Date: Mon, 24 Jan 2000 15:15:26 -0500 From: "Michael Rothenberg" <rothenberg@automationonline.com> To: "Nik Clayton" <nik@freebsd.org>, <questions@freebsd.org> Subject: Re: ipfw, multiple ISDN TAs, munging routes automagically Message-ID: <00f601bf66a7$c6af1dc0$3301a8c0@ias.com> References: <20000124193455.A55129@catkin.nothing-going-on.org>
next in thread | previous in thread | raw e-mail | index | archive | help
ER... hehehe sounds fine except for the following: you have to be able to split your net in half at a particular address. each 'half' would need to be running the same sum load. so you really need to get the load from each internal source then resplit based on load sums of the sources in each 'half'. problem is if one host in the middle of your address table is sucking up a whole bunch of your BW then you will have to make more then 2 'halfs' or groups, no? Might be easy if all you have are workstations. Might be hard if you have a few servers interspersed in there. But I dont see anything in ipfw(8) either that would prevent this. Just my guessing here. No knowledge behind it };) Michael, the BSD newbie ----- Original Message ----- From: "Nik Clayton" <nik@freebsd.org> To: <questions@freebsd.org> Sent: Monday, January 24, 2000 2:34 PM Subject: ipfw, multiple ISDN TAs, munging routes automagically | Hi folks, | | I *think* ipfw can do this. But before it to a client, can someone | confirm for me whether or not the following is possible. | | Consider a network, with a FreeBSD (probably -stable, but I can use | -current if absolutely necessary). The FreeBSD host has 3 interfaces; | a regular ethernet interface, and two ISDN terminal adapters, both | doing ISDN B channel bonding, for a total of 256 Kbps. | | Now, what we want is for one half of the external network traffic | to automatically go up one of the ISDN TAs, and the other half to | go out of the other TA. Each TA will have a different IP address | assigned to it. | | I don't think I can do this with regular routing. Correct me if I'm | wrong on this, but I'm pretty certain about it. | | So I've got the following evil plan in mind. | | Run ipfw on the FreeBSD machine, with 2 rules. Both rules look at the | source address of the packet. If the source address is in the first 50% | of addresses on the internal network, then use a 'fwd' ipfw rule to | forward the packet on to the first ISDN TA. If the source address is in | the second half of the internal network, use a 'fwd' rule to send it out | to the second TA. | | Oh yeah, and I'll be doing NAT at the same time. | | When you've finished retching from this gross hack (which will only be | used for a couple of weeks until the fibre is laid), can anyone confirm | or deny whether or not the above will actually work? I don't see | anything in ipfw(8) which would preclude it, but I thought I'd check | just in case. | | Cheers, | | N | -- | If you want to imagine the future, imagine a tennis shoe stamping | on a penguin's face forever. | --- with apologies to George Orwell | | | To Unsubscribe: send mail to majordomo@FreeBSD.org | with "unsubscribe freebsd-questions" in the body of the message | To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?00f601bf66a7$c6af1dc0$3301a8c0>