Ways of creating a WAN

R

Rees

n00b
Joined
Sep 23, 2004
Messages
9
I have a project where there are going to be 6 nodes on the network and they are 20 miles apart from each other... I was thinking should it be a Local Area Network or a Wide Area Network?

Does anybody know how this can be done?
P.S This project is Simulated...

Thanks
Rees
 
You can try a VPN( Virtual Private Network) Put all the node in to the VPN and it will act like one big network. Over the Internet of course.
 
If im right it would still be a LAN, as the network is confined to a gerographicly confined area.

But then there is more than just LAN and WAN to consider.

As for how to do it is beyond me! i dont of any cables that could run 20miles, you would be looking at loads of repeaters and thats not possibly (beyond 4 repeaters i think)

Why are they 20 miles apart? :confused:
 
A VPN runs over the Internet so lets say me and one wanted to make a VPN im in Canada and Your in England. We would use VPN Software or i think some routers have it built-in but we would be using the internet as a connection Oh and in my mind 20 miles apart is still a LAN, i perfer a WAN defintion to be like 100s of miles apart from me to you for example
 
Um lets me think if im not mistaken max distance of cat5e cable is 150m whichs is what in feet? im not about 450 feet so more then 4 would be needed
 
Also if this is simulated and not costing anything why not go fiberoptics it can go for some distance but it still might need a boost before the 20 mils is up.... hmmmmm im not sure about distance on fibre but hey if its only simulated why not go fibre
 
if there 20 miles apart from each othere then this would sort of me a circle shape why not set up a router in the middle of all 6 and have fibre running to that and from the router to all 6 that would be cool... I like when things are "simulated" cause i have lots of ideas but when it comes to actual then i have to think about cost and stuff like that
 
WOW! didnt expect so may replys

The project is simulated but all the users needs are effected by which transmition media/meidum I use... I will think about these ways you have suggested...

how much would it cost to run optical fibre 20 miles?

Thanks
Rees
 
I think the best way to do this would be either full or fractional T1s (or bigger, depending how big these sites are supposed to be) connecting to once central location.

A secondary option would be vpn connections to one central site. Most WAN connections are point to point, so having them connect in to 1 central site provides the most efficient network design.

Running fiber would be prohibitively expensive. Best to leave that sort of thing to the telcos.

How many users are supposed to be at these sites?
 
Rees said:
I have a project where there are going to be 6 nodes on the network and they are 20 miles apart from each other... I was thinking should it be a Local Area Network or a Wide Area Network?

Does anybody know how this can be done?
P.S This project is Simulated...

Thanks
Rees
Click to expand...

You're missing information. Or is this all that you're given? If you are to create a LAN environment you would need single mode fibre to connect these six sites. Of course that would entail laying and burying the fibre but that part of it might be beyond the scope of this discussion.

Otherwise you should go with something more cost effective than full T1s. Frame Relay, sometimes referred to as fractional T1s because you don't always get the full speed of a T1, would be the way to go. T1 lines are billed "by the mile" where Frame Relay is a flat fee (relatively) based on bandwidth.

In the long run single mode fiber would be cheaper as you would own the lines and get great bandwidth out of it. Essentially you could be your own ISP and sell off your services at that point. :p :D

Good luck with your project!
 
The question has a pretty big hole in it.. What exactly will they be doing? If it's just to access an accounting system, or for parts, or alike you could use modems.. If they are using something more intensive (higher bandwidth requirements) then as others have suggested you would want to look into Frame Relay. VPN would be your last choice as there are far too many variables involved, that is unless their connections aren't critical.
 
Rees said:
WOW! didnt expect so may replys

The project is simulated but all the users needs are effected by which transmition media/meidum I use... I will think about these ways you have suggested...

how much would it cost to run optical fibre 20 miles?

Thanks
Rees
Click to expand...

More than you have. Considering a 3m cable is like $70.
 
There are a lot of factors to this and here are some of the things to consider.

The most expensive route initially is going to be to set up your own fiber. You are going to have to not only buy the fiber and fiber routing equipment but you are going to have to actually lay the fiber which is a massive construction cost as it will likely entail tearing up other people's property and public streets to do so. Once it is up and running, this option will provide the best overall throughput but it is the most succeptible to the "Backhoe Effect".

The next option is a frame relay network, commonly called point to point. This may or may not work for your application as these networks are frequently not as private as you would expect them to be. Most telco's run multiple point to point systems over the same equipment which is not secure enough for certain things like banking applications. You may or may not have similar regulatory issues to deal with if you go this route. You can get good throughput and even managed services with this method but you can rest assured that the telco is going to charge you a premium for going this route not only for setup but monthly.

VPN would likely be the second most expensive initial deployment but would be the most cost effective over time. It also has the capability of being the most redundant of the three primary options because you can utilize multiple cheap internet access lines at each site and mesh the whole shebang using stateful failover and intelligent routing.

Personally, I would really consider the VPN option. It offers the most flexibility and expandability for the best price point. In addition, the security that you can implement is superior to the other methods. Private fiber used to be more secure but a method to snoop fiber wihtout T-tapping was recently developed.

-Vaporware
 
joecool234 said:
More than you have. Considering a 3m cable is like $70.
Click to expand...

You lease the fiber from a telco.. The guy couldn't run it himself even if he wanted to hehe ;)

Vaporware said:
Personally, I would really consider the VPN option. It offers the most flexibility and expandability for the best price point. In addition, the security that you can implement is superior to the other methods. Private fiber used to be more secure but a method to snoop fiber wihtout T-tapping was recently developed.

-Vaporware
Click to expand...

The only thing I don't like about VPN is you can't have any kind of SLA simply due to the fact that any 8yr old script kiddie can ddos that remote right off of the network. It's better to have a frame where you can call the telco and say "it doesn't work fix it now!" rather than "Please Mr. ZeraximusPrime stop ddosing us, I promise I'll buy for you next time I see you at the local Piglies"... Just my humble opinion ;)
 
I would have to agree with the VPN route. Fiber would be faster, but it is quite expensive for the initial investment and long term usage. As stated above, VPN does require a internet connection at each location but that is a miniscule cost considering the equipment requirements and labor required to run fibre.
 
Frontline said:
Um lets me think if im not mistaken max distance of cat5e cable is 150m whichs is what in feet? im not about 450 feet so more then 4 would be needed
Click to expand...


"The TIA/EIA-568 standards or the international ISO/IEC standards for generic structured wiring have not changed the basic dimensions of the horizontal cabling system. Beginning with the first issue of these standards, the horizontal link has been defined as (1) a straight run of maximum 90 meters (294 feet) of solid core copper cabling from the interconnect or cross-connect in the telecommunication room to the outlet in the work area and (2) a set of patch cords to connect the end-user equipment in the work area as well as patch cords in the telecommunication room. The length of the horizontal solid core cable run should not exceed 90 meters or 295 feet. The total length of all patch cords and equipment cords should not exceed 10 meters (33 feet) such that the channel or the end-to-end connection between the network devices does not exceed 100 meters (328 feet). These definition are independent of the media type."
 
Before anybody goes any farther with this let's clear up this question of LAN vs WAN.

A Network confined to a building or set of buildings in a campus is the generally accepted definition of LAN. Most Network Analysts (such as myself) would even go as far as to say that if you have right of way to install fiber, then it would still be a LAN if it is within 2 Kilometers. This is the distance limitation of most consumer fiber optic products. Over this distance and you are talking WAN regardless of the fact that it is a private network.

Also, I have worked on Fiber projects, Unless you can afford close to $1million(US)/mile then don't even think of trying to run your own fiber.

Now, You haven't given enough input to get the correct data from your Simulation::

info still needed::

What is the projected bandwidth usage? - this determines what size/type of pipes you will need - ranging from always-on modems to DSL/Cable to frac T1/T1(and up) to Wireless.

What services are available in the region? - this determines what size/type of pipes you will actually be able to get - many rural areas still don't have any High Bandwidth facilities available at a reasonable cost.

Time/Cost projections - all simulations have to have limits, and this will be your biggest, is this to be a low-cost long-term migration? or an any-cost mission-critical immediate upgrade?

What type of security infrusructure is required? - This makes a huge impact on the overall cost of the project because of all the extra hardware and planning required to meet NSA/DOD level specs.

What type of redundancy is required? - This also makes a large overall impact on the project as full redundancy would double the cost of the project and the maintenance.



Of course, each of these questions then leads to several sub-questions that will build the framework of what you need to look for and implement. The rest you can do on your own, taking blind advice on buythis/build that always hurts in the end.
 
The things you need to look at first and identify are this:

1 How much bandwidth is going to be needed. - This is based on what will be done at each site

2 Are the users just browsing the internet or are they streaming live p0rn shows.

3 What apps would they be using - would a thin client front end a server farm on the backend work for whatever they are running? I.E. - you can have users dial up to the internet or provide cable/DSL connection that they share and they would then launch a thin client like Citrix and connect back to a server farm where your published app is. You can even use SSL via VPN and use Citrix through Internet Explorer.

4 How much loot do you have to cover the costs - a full T1 runs my company around $1,500 a month - a frac T1 (frame relay) is less depending on how much our port speed is and our CIR for each frac T1.

5 VPN is very nice but for more then 5 users you need a concentrator to be able to handle the requests and traffic. Don’t think a Linksys box for $80 would cut it.
 
Rees said:
WOW! didnt expect so may replys

The project is simulated but all the users needs are effected by which transmition media/meidum I use... I will think about these ways you have suggested...

how much would it cost to run optical fibre 20 miles?

Thanks
Rees
Click to expand...

Digging costs + equipment = One Shitload (yes that is a real technical term)

I supervised a project in which we dug fiber on a sonet for about a total of 8 miles. Cost = 10.8 Million US...

Nuf said there.
 
xaruum said:
info still needed::

What is the projected bandwidth usage? - this determines what size/type of pipes you will need - ranging from always-on modems to DSL/Cable to frac T1/T1(and up) to Wireless.

What services are available in the region? - this determines what size/type of pipes you will actually be able to get - many rural areas still don't have any High Bandwidth facilities available at a reasonable cost.

Time/Cost projections - all simulations have to have limits, and this will be your biggest, is this to be a low-cost long-term migration? or an any-cost mission-critical immediate upgrade?

What type of security infrusructure is required? - This makes a huge impact on the overall cost of the project because of all the extra hardware and planning required to meet NSA/DOD level specs.

What type of redundancy is required? - This also makes a large overall impact on the project as full redundancy would double the cost of the project and the maintenance.



Of course, each of these questions then leads to several sub-questions that will build the framework of what you need to look for and implement. The rest you can do on your own, taking blind advice on buythis/build that always hurts in the end.
Click to expand...

Yup _ have more or less the same question on this one as well...
 
At the telco I work for we are working on deploying our own fiber. 1 mile of fiber ran aerially is right about $7,500. Burried it was well over $15,000 a mile.

Rees said:
WOW! didnt expect so may replys

The project is simulated but all the users needs are effected by which transmition media/meidum I use... I will think about these ways you have suggested...

how much would it cost to run optical fibre 20 miles?

Thanks
Rees
Click to expand...
 
You must log in or register to reply here.
Back
Top