What module card should I get for Cisco 1721?

school, we used Cisco smart cable to connect two routers to act as WAN link. Most Cisco 1721 on Ebay.com don't have this serial interface module cards. They usually have a T1 DSU/CSU module card. Can this T1 connection function as WAN link? Or should I get maybe one serial interface module cards for each 1721 router?

Either card would be fine. This model router is EOL'd, so anything should be cheap on the surplus market.

A WIC-2T card (which you mention first) goes for ~$25 on eBay. a 1.5ft smart serial DCE/DTE cable goes for $10.

A WIC-1DSU-T1 card goes for $5-$15 or so. A Crossover T1 cable goes for $4-$5 or so. (or you can make one).

Either the Synch serial card or the T1 card will function near identical for you. With the T1 card, you'd get a little bit more to what goes on in "real life" for WAN circuits.

Yes, you can hook up two WIC-1DSU-T1 cards with a cross-over T1 cable.

Hi Doug, Thanks for your answer. Is the cross-over T1 cable you mentioned the same as a cross-over cable used to connect a host to a host?

Another question. I know you said either WIC is good. But I can think of a situation where a WIC-2T is preferred than WIC-1DSU-T1. There are lab situation a router functions as a frame-realy switch, the WAN connection from two branch offices takes up two ports, and another ISP WAN connection takes up the third connection. A 1721 router with 2 WIC-2T would offer 4 serial ports, enough for this prototype. While a 1721 router with 2 WIC-1DSU-T1 offers only 2 ports. And I can't add another WIC-1DSU-T1. For reasons like this, should I go for WIC-2T?

Thanks.

same as a cross-over cable used to connect a host to a host?

No, not the same pinout. Same idea, same function, different pins.

situation where a WIC-2T is preferred than WIC-1DSU-T1. There are lab situation a router functions as a frame-realy switch, the WAN connection from two branch offices takes up two ports, and another ISP WAN connection takes up the third connection. A 1721 router with 2 WIC-2T would offer 4 serial ports, enough for this prototype. While a 1721 router with 2 WIC-1DSU-T1 offers only 2 ports. And I can't add another WIC-1DSU-T1. For reasons like this, should I go for WIC-2T?

There are also 2 port T1 cards (and while not available on the 1721, also quad port T1 cards on other routers). The VWIC-2MFT-T1/E1 are readily available used as well, since they only work in old hardware.

Again, both would function well, its just that you would be much more likely to encounter a WIC-1DSU-T1 or VWIC-2MFT-T1/E1 in "real" life more often. Ie. out in the field, I only saw a couple installations using 1721 and WIC-2Ts. By far, whenever I had a synch serial port hookup, it was with even earlier hardware (ie. 2500, 7000/7500). Once the integrated CSU/DSU version was readily available, the market switched almost instantly.

One benefit of the WIC-2T vs. VWIC-2MFT-T1/E1 is the IOS revs supporting the WIC-2T go back to the start, while the VWIC part came out much later, probably midway through its lifespan. The VWICs aren't as readily available as the WIC-2Ts, but are still there.

I should followup why that would be important. If you need a newer IOS that you have, you may also have to upgrade the DRAM and Flash in the router in order to run the code rev you need for this card. (as well as some how obtaining the version of code you want in the first place with zero support available from cisco..).

And by that time of course you have spent more than if you would have bought an uptodate router from the 800 series...

Is it expensive to upgrade the DRAM or Flash? DRAM is quite cheap for PCs. Flash memory is a little more expensive. Do Cisco routers use special expensive DRAM or flash?

The newest IOS version is 15.0. Is it likely for me to upgrade 1721 to version

15.0?

With WIC-2T and a smart cable, I have to issue 'clock speed 64000' on one end of the connection. Do I have to do the save when using WIC-1DSU-T1 or VWIC-2MFT-T1/E1?

And how would you get an 8xx series router to test out T1s or Frame-Relay switching? Ethernet is not the only technology.. (ignoring the xDSL options in that line which doesn't help out either).

Well it depends on what you want to build your testlab for, of course.

T1 (E1) and framerelay is completely dead here. Everyone uses DSL and ethernet. Maybe it is different in his part of the world, but I want to warn him for spending money on obsolete stuff like 1721s and WIC cards, only to find that he needs to spend more money on overpriced memory expansions and a license to get a recent IOS, and still have a device with obsolete performance.

You can find plenty of options for upgrades on eBay as well if you happen to get a router that needs more. Or there are stores that specialize in odd upgrades like this. I've used MemoryX often over the years to get cisco memory, although their prices will be slightly higher than eBay. Ie.

As with anything, it depends on what is available, if it needed a module that flooded the market and used everywhere, it'll be dirt cheap. If it needs a module that was fairly rare and expensive (ie. linear flash memory for certain router lines instead of ata), it could get pretty expensive.

The router platforms used quite a variety, some items were pretty readily available and cheap, some items were very rare and only used on Cisco.

No, the release of 15.x cleared out and stopped development on a lot of legacy hardware like the 17xx platforms. You can run 12.4/12.4T on them given enough DRAM/Flash. Use the FN to find out available versions at

There really isn't much need for you to know 15.x IOS vs. even 12.2 IOS. First off, 15.x was mainly a realigning of platforms and clearing out of older gear. Also, the basics are the same going back quite a long ways. The newer revs, I'd say introduce more advanced, specialized features..

You don't set the clock speed like in a synch line, but you would have to set one end to generate clocking and have the other end sync of line clocking (which is default). Should be easy to find this out there.

T1 and framerelay is dead? It took me a few days to read the frame relay chapter in my CCNA book. I'm actually in New York. Then what's the current technology when it comes to WAN?

Well, *I* still have T1 customers as a service provider. If that is where you want to end up, knowing how to support legacy WAN circuits would be a big plus.

I wouldn't expect to be installing *new* T1 data circuits though. Installing new T1 voice circuits (specificly in the form of PRI) would be an active area though. Troubleshooting T1 circuits in general, and specificly for voice services (although fundementals of T1 are the same for both voice & data with tiny tweaks either way) would be again a plus.

WAN areas are taken over by optical 10G for service provider, MOE for larger customers, and probably some form of DSL and/or cable modem for smaller customers. Most likely with ethernet handoff. Knowing PPPoE protocol over either would be a benefit I would think.

(no mention of frame relay in there, is there. :( At least in my area, the ILEC priced frame-relay way out of proportion to ptp T1 links long ago. Kind of funny as they built up a lot of service gear behind their cell-relay/frame-relay integration, but then charged a boatload for it driving advanced customers to ptp t1 long ago.

I agree with that. Also add the configuration of VLAN. Providers often use different VLANs for different type of service here. (and then PPPoE on top of the VLAN subinterface)

It is difficult to build your own personal standalone testlab because the network is not peer-to-peer. A DSL router talks to a DSLAM and the PPPoE client talks to a PPPoE server, and you need to either have one or preferably two connections at a provider, or you need all that equipment at home to have a usable testlab.

Connecting two routers over a crossed cable provides a lot of insight but is not a setup you encounter in reality.

Wow! Do you experts have a suggestion on the best way to get into networking field? So that I don't try CCNA for one or two years and realize it's working for me job-wise and have to go a different direction.

I'm not the best one to give career advice, because I don't work in a typical environment.

To me though, the CCNA, like a degree, represents that you can concentrate enough to complete a program, and you should have the fundementals down pat. I'd think that an employer looking for a position would chuck those that don't have at least a CCNA or equivilent without consideration. Ie. you'd have to really shine out and be outstanding if you didn't have the basics and paperwork with it.

The point of the CCNA isn't to show that you can setup a frame-relay switched network, but more that you have learned the fundementals. Ie. from what I remember of the CCNA program, you would really get to know IP subneting (ie. know how many IPs are in a /27, in a /22, in a /18). Can subnet any given block. Can setup a network device to route say a /27 to something else. Know what VLANs are, how to use vlans on a switch. Do L3 routing on a switch. Know how WANs are normally setup, how T1s are typically deployed, even if they are dying breed, I still support my T1 customers. They do add on more T1s to their PPP ML bundles. These are all still very much in use (unlike Frame-Relay switching).

Knowing the fundementals is more important than the details. The ability to learn, and keep learning all the time is important. The goal isn't the certificate, it is to know the basics down pat to go to the next level. Once you know the fundementals, and have the ability to learn, it does get to be easier. Knowing the test answers for rote regurgiation isn't going to get you as far as really knowing it.