This one is just too juicy to pass up :-) The concise question is "Why would you use linux/freebsd or an appliance based on them instead of purchasing a real router?"
Cisco rules the backbone, they do enough on lan switching to get my vote their, and their Aironet radios are OK if you're abusing 802.11b by running it as a wireless internet provider but ... CISCO KNOWS SQUAT ABOUT ROUTER/WIRELESS INTEGRATION! There, I've said it. Detaled proof of concept is below: Lets consider a network I recently installed. I've got a tower on a building on a bluff that overlooks a valley where there is no DSL service. On the opposite side of the valley I have another tower on a building on a bluff. The two towers are separated by about three miles. Some sites in the valley can see one side, some can see the other, neither site would cover the whole valley properly. The building closest to my head end has an elevator shaft on top with my small tower bolted to the side. There is room for a good sized 19" cabinet and AC power. The building on the opposite side of the valley has a beautiful 50' tower on top but the only facilties on the roof are a NEMA outdoor enclosure. The NEMA enclosure can take a couple of radios and maybe a small hub but would not take any router larger than a 1605. So, this one building has a backhaul link, it has a customer cell, it feeds the cell across the valley, and I have two other customers attached to this location that received dedicated point to point links for a total of one cell and four point to point wireless links. There is also a wired link to our customers in the building. From the layer three perspective this location has four /29s used for point to point (why not /30s? you tell me, but ponder the wonder of VLANs and OSPF first :-] ), one /27 that is the customer cell, and a /29 for the wired customers. How would you build a network with six total subnets, one wireless cell, and four point to point links, using only Cisco equipment? This is the parts list with estimated pricing: Cisco 2620 $1500 Catalyst 1912 $300 Aironet AP352 $650 four x WGB352 $1800 total $4250 Now consider the alternative - a $400 PC, a $50 flash card, $200 for a linux based wireless capable OS from http://www.mikrotik.com, two PCI Aironet cards ($350) and a Dlink quad port ethernet card ($100). $1100 total I still needed the $650 AP and one external $450 WGB352 - $1100 for this. So, for $2200 I got the effect of purchasing $4250 of refurbished Cisco gear. Now lets look at the soft benefits of MikroTik vs IOS: Feature MikroTik IOS ssh yes with provider image OSPF yes yes PPTP server yes not that I can find PPPoE server yes yes, clumsy rate shape PPPoE client bit rate per user nope, just IOS traffic controls NAT yes yes processor AMD 1 gig 40 MHz PPC memory 256 meg max out at 64 meg firewall ipchains(cool) clumsy access lists user interface killer CLI clumsy CLI annual support $500 or so including aironet about 3X MT costs Over a three year period the MikroTik box is going to cost me about one fourth of what I'd pay for a minimal Cisco solution and the performance of the MT is on par with a 7206 /w an NPE300. Now there are a lot of reasons to prefer Cisco in the backbone over MikroTik, but I'm not going to say anything :-) You'll just have to download the free eval of MikroTik and make it work with your Cisco lab if you want to know ... > I always see people asking for Linux/Unix sims for BGP > or just to simulate a router in General. I for the life of me can't > figure out why a Linux router could be better than a 2501 and > it would cost me more to build a Linux box. I see the same with Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=46107&t=46107 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
