If you are looking at a very small scale SLB solution, why not just use IP SLB directly on the MSFC? It can handle quite a few sessions/pps, and if you feel overload, you can upgrade to Sup720-3B, which has HW acceleration for IP SLB.
If you stop using destination NAT, it would be the perfect solution for you. If you need some assistance with configs, unicast me. Here is the link for the IP SLB docs: http://www.cisco.com/univercd/cc/td/doc/product/software/ios122/122newft /122limit/122sx/122sxf18/slbsxf1.htm Arie -----Original Message----- From: Paul Stewart [mailto:[EMAIL PROTECTED] Sent: Wednesday, July 11, 2007 19:41 PM To: Arie Vayner (avayner); 'Robert Blayzor' Cc: [email protected] Subject: RE: [c-nsp] SLB Question Thanks Arie... a quick look at the pricing makes this option prohibitive ... nice option though...;) If this was for a large scale rollout then the ACE would be ideal.... This is pretty much just for two servers which are mirrored/redundant to one another - looks like doing it on the server level is going to be best option or go with something like Barracuda's load balancer appliance (can get a pair of them redundant for under $10k) but we like to standardize on Cisco as much as possible... Thanks Arie and everyone for your input as usual.... Paul -----Original Message----- From: Arie Vayner (avayner) [mailto:[EMAIL PROTECTED] Sent: Wednesday, July 11, 2007 12:30 PM To: Robert Blayzor; Paul Stewart Cc: [email protected] Subject: RE: [c-nsp] SLB Question Paul, I recommend that you look at the ACE module. It's the latest and greatest HW based SLB solution (a blade on the 6500): http://www.cisco.com/en/US/products/ps6906/index.html If you really need to be able to provide access to the VIP from the same L2 domain where the real servers are located, I would recommend moving to dispatched mode. Just disable "nat server", and configure the VIP as a loopback interface on the real servers (easily done on any OS). This would allow the servers to use the VIP as a source, so both remote and local clients would be able to access the VIP on the SLB instance, but the return traffic would not require a translation back... Arie -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Robert Blayzor Sent: Wednesday, July 11, 2007 19:02 PM To: Paul Stewart Cc: [email protected] Subject: Re: [c-nsp] SLB Question Paul Stewart wrote: > So, what's a good hardware solution (bearing in mind that we can still > do this in software on the servers)?? Cisco used to make load > balancing hardware at one time but I don't think they are involved > with that any longer?? Open to hardware suggestions.. preferably > something that works transparently as a bridge and can work as a pair > of hardware devices for fault tolerance within the hardware? Cisco still makes some of the CSS line, but not sure that fits the bill. I'd actually take a look at the Foundry ServerIrons. If you want a good laugh, the Foundry ServerIrons are more IOS like than Cisco's own CSS's. ;-) The ServerIrons are solid and are quite reasonably priced... vs what's left in the CSS line. (I believe the CSS 11501's are EoSale). -Robert _______________________________________________ cisco-nsp mailing list [email protected] https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/ _______________________________________________ cisco-nsp mailing list [email protected] https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
