[Webinar] Streamline your web hosting managementRegister Today

x
  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 2109
  • Last Modified:

Leased line/network topology scenario

Hi there

I am studying for one of the Cisco modules, and trying to get a scenario correct in my head.

Say an HQ had 10 branch offices. If these were connected to each other with Cisco 2600 routers on a leased line (using HDLC in a hub and spoke layout with EIGRP as the protocol), how would the branch offices connect to the internet?
Could the branch routers have an additional interface to an ISP's network? Or would the HQ handle the internet, in which case although this would be less expensive, it's a lot of strain on the lines...

And seeing as the 2600 router is designed for branch office use, what would be recommend for HQ (say HQ had 100 odd users, hosted the Exchange servers, and each branch office had about 30 users).

Any opinions?
0
Dilan77
Asked:
Dilan77
  • 7
  • 6
1 Solution
 
Don JohnstonInstructorCommented:
>Could the branch routers have an additional interface to an ISP's network?
Yes.

>Or would the HQ handle the internet
Yes.

>in which case although this would be less expensive, it's a lot of strain on the lines...
Correct.

There's really no "right" answer. Both solutions will work. The question is, which one does the Cisco exam expect you to select?

>what would be recommend for HQ
It's not really about the # of users. It's more about the amount of traffic. What kind of leased lines? T-1? At 1.5mbps, that gives you about 45mbps. Assuming max utilization, that would require a minimum of a 3600. If the leased lines were slower...

http://www.cisco.com/warp/public/765/tools/quickreference/routerperformance.pdf

But these certification design questions are difficult to answer unless you have the courseware to refer to.

0
 
Dilan77Author Commented:
>There's really no "right" answer. Both solutions will work. The question is, which one does the Cisco exam expect you to select?

Probably the most expensive one ;)

Only joking. This isn't an exam question as such...it's just that when I've been studying for the CCNA and BCMSN, there are always example networks in the books that the authors chuck in to give you an idea of a real life scenario. A  friend is in a situation where his branch-HQ connections are currently via VPN, but they're looking to implement leased lines, so I was using that as a real world example.

Let's say the traffic will be mainly web and email, nothing fancy like video. If HQ has a leased line of 1.5Mbps, that would need a 3600. If each branch used a fractional T1 (or E1 seeing as we're in Europe) of 1Mbs then that would require a 265x, correct?

0
 
Don JohnstonInstructorCommented:
>Probably the most expensive one ;)

Actually, in the CCDA exams, the right answer usually is the most expensive one. :-)

The easy answer is to add up the bandwidth of all the working interfaces and then select a platform that can handle that amount.

The hard answer is to consider what is the maximum amout of traffic the device will see at a given time. Since it's unlikely that every interface will be at max utilization at the same time, the actual utilization would be significantly less.
0
The new generation of project management tools

With monday.com’s project management tool, you can see what everyone on your team is working in a single glance. Its intuitive dashboards are customizable, so you can create systems that work for you.

 
Dilan77Author Commented:
Assuming he went with the HQ providing internet connection to the branch offices via the leased lines and NAT was carried out by a firewall at HQ before traffic went to the internet, would anything else need to be configured into the routers at the branch office, or just the usual?

Assuming the branch office's local subnet is 192.168.10.0/24 and all the interfaces on the branch router (the fa0/0 local interface and serial WAN interface) have been assigned the correct IP and subnet mask.

router eigrp 10
network 192.168.10.0
network 10.0.0.0 [classful address being used on WAN links, even though a /30 is used between sites]

HDLC would be the WAN encapsulation which is default anyway.

0
 
Don JohnstonInstructorCommented:
That's about it. Assuming the HQ router is advertising a default route to all the branches.
0
 
Dilan77Author Commented:
Thanks....

Let's say we had one router at HQ that had an interface to the ISP and also the 10 serial interfaces where the leased lines terminated. It would be using HSRP for failover.

The network between the ISP and the relevant interface on the router was 62.180.27.0

Am I correct in thinking that the best way to use a default route to the internet would be

ip default-network 62.180.27.0

Since, this way, other routers in the WAN receive this route to the internet automatically?

Would HSRP be a good failover method in this case, or is there another one more commonly used? If we go with the one router scenario, and there was no failover, then that would cause a big problem were that router to go down...
0
 
Don JohnstonInstructorCommented:
>Let's say we had one router at HQ that had an interface to the ISP and also the 10 serial interfaces where the leased lines terminated. It would be using HSRP for failover.

HSRP and "one router" are mutually exclusive. :-)

In order to use HSRP, you have to have at least two routers.

When you have one router (or path), you have no failover of any type.

What I would do is to have a default route on the HQ router and redistribute that with EIGRP to the remote sites.
0
 
Dilan77Author Commented:
Hi

Sorry, didn't mention the second (failover) router that would be used with HSRP. :)

Would this (the two router HSRP) model provide the necessary redundancy or would the redistribution method be better? The first is more costly, but I'm wondering if HSRP would actually work in a situation such as this?
0
 
Don JohnstonInstructorCommented:
No. HSRP is only used to provide default gateway redundancy for end stations.

For example, say your workstation is on a network that has two routers connected to it. Both routers have a path to the outside world. You use router 1 as your default gateway. If it fails, you would need to reconfigure the default gateway on your PC to use router 2 in order to get outside. With HSRP, this would failover automatically.
0
 
Dilan77Author Commented:
Ok, thanks (and final question :) )...what would you as a failover should the HQ router go down and seeing as it was the route to the internet and the spoke of the leased lines?
0
 
Don JohnstonInstructorCommented:
How is the second router connected to the network and what else is it connected to?

Keep in mind that the router is but one of numerous single points of failure.

There's the connection to the internet, the connection to the remote sites and the router itself. In order to be truely redundant, you'd need to connections from each remote site. One going to Router 1 and the other going to Router 2. Then each router would need it's own internet connection. Once that's done, the routing protocol would handle any failures.

Many businesses can't justify the expense of this type of network. So they use DDR (dial on demand routing) or low-cost internet connections at each site or hot spares or any of a number of other methods of keeping traffic flowing.
0
 
Dilan77Author Commented:
The second router isn't connected to anything, it's a hot spare in case of hardware issues on the first.

I guess an ISDN backup at each site would be worth it in case the link or HQ router did go down.

In any case, I've managed to go wildly off tangent ;)

Many thanks for your help :)
0
 
Don JohnstonInstructorCommented:
Design issues are always tough. Sometimes there are just no "right" answers. :-)
0

Featured Post

Take Control of Web Hosting For Your Clients

As a web developer or IT admin, successfully managing multiple client accounts can be challenging. In this webinar we will look at the tools provided by Media Temple and Plesk to make managing your clients’ hosting easier.

  • 7
  • 6
Tackle projects and never again get stuck behind a technical roadblock.
Join Now