Creating a CAN

[BobSpark1999]

Hi all,

I am designing a Campus Area Network as part of a piece of homework.

I am a beginner in networking, and finding resources about specific is rather difficult.

I currently have planned to have a DMZ containing a proxy server (to access web browser through), a FTPS server, a web server, and a VoIP server. Then on the main network, I will also have a DNS Server, Print Server, DHCP Server, DFS File Server and a database server.

My question:

DNS Server, by my understanding, converts domain names into IP addresses and vice versa - similar to looking up in an address book. The Domain controller controls the user accounts for students etc. Therefore, do I also need a domain controller on my network, aswell as a DNS server?

Secondly, servers themselves. I have googled "How much does a DHCP server cost" so many times etc. and I can't seem to find a simple answer. I come the the conclusion its all one piece of software (Microsoft Server 2008 is what I've found) and you configure the software. So my next question from that; If I buy the disc for Microsoft Server 2008 to configure say, a DHCP server.. Do I then need to buy it again to configure the next server (a File server)? Or, can I use the disk to install the software again? I.e. is it a 1-time-buy for unlimited software installations? My second question to that, what is the "physical device" called on which I deploy my configured server, and presumably, I could buy 10 of them (one for each server)?

I'm aiming to have an all microsoft network.

I have asked in many places, and my question keeps being shut down as "too broad".... please can someone help me out here!

Cheers,
Bob

[Philip]

Seems you will be attempting to build a complex network environment without the proper knowledge - may I suggest hiring network professionals with proper background/education to build it correctly, or at least in advisory capacity?

To answer some of your specific questions:

DNS servers are used to translate domains/hostnames to IP adresses, yes.
DHCP servers lease out dynamic IP addresses to clients.

Both those types of servers (and other types, i.e. DC, etc.) can be on the same physical machine, they are software packages. DHCP is often off-sourced to some type of appliance, like a firewall or a NAT router. In other words, a DHCP server can be a function of your NAT router, of your Firewall, of your DC, of any server, etc. It is up to how the network is set.

Properly configuring a DNS server, a domain controller, file servers, etc. in a large campus area network is not trivial, it cannot be accomplished by simply buying a separate "server" computer preconfigured for each task.

[BobSpark1999]

Hi Philip,

May I just remind you this is a piece of homework, I'm not actually building the network. But I am trying to learn about networking so I understand this much better. The way my lecturer put it is that each 'server' package is a piece of software, and each piece of software can be deployed on separate devices.

I am aware you don't buy them preconfigured, that's not at all what I asked. I'm also aware of what DHCP servers do, but my question is
- What is the name of the physical device/machine the software server is deployed on to?
- Do I need to buy 1 piece of software per server I configure, or can I use 1 peice of software to deploy multiple servers?

[Philip]

Oh, I missed the "homework" part.

A DHCP/DNS server can run on a computer (under Linux or Windows), or on an "appliance" like a router/firewall device running some type of mini-Linux version.

Some types of server packages (DNS / DC) can be deployed over multiple physical machines as well and communicate with each other. There is no 1:1 correlation between software "server" packages (DHCP, DNS, DC) and physical servers. Also, some software packages and operating systems are free/open source, so you don't necessarily need to "buy" them.

[YeOldeStonecat]

A Campus Area Network is just a larger network with segments to it, typically spread across several buildings. Picture a college campus. Or a business that has several buildings next to each other, across the street, etc. Sort of like a wide area network..but...close together.

In the old days we connected the buildings usually with ethernet or fiber trenched underground through conduit. These days we often do it more and more with point to point wireless radios...not wifi..but radios such as Ubiquiti airMax or airFiber..designed for very high speed point to point connections. Basically it makes an invisible ethernet cable through the air.

Depends in the size..and amount of "nodes"...different buildings may have their own router creating a differenet subnet for each building. Can have DHCP and DNS running for each building.

A Windows Server ...running active directory (a domain) is usually more for a controlled office network. It controls the users and devices on the network. It runs DHCP and DNS for those workstations..but often on a campus area network you have many more devices that don't need to connect to that office/production network..they're seen as guests and just need to access the internet, so you have other devices such as routers/firewalls running DHCP/DNS for them.

[BobSpark1999]

Hi,

Thank you for the replies!

Okay, so that makes a bit more sense. Just to clarify some things. Lets say I have 5 buildings. I've designed it such that I have:
* Note B1 = Building 1

Internet > (B1) Router > Firewall > Campus Switch (L3)
From there, the Campus switch connects to 10 L3 switches (2 per building - 5 sets of Aggregation switches) and a further L3 switch for the DMZ (containing web server).
Off these aggregation switches each come 4-5 more L2 switches (depending on size of buidling/number of floors, generally 1 per floor).
Off each L2 switch is say 30 fixed PCs and 2 printers.

Am I correct in thinking that the 2 aggregation switches (L3) per building are creating 'subnets' in each building.

How do the IP addresses work with regards to aggregation switches?
If my building 1 subnet was 192.168.5.0, does each L2 switch have its own IP too? (Does it need one? I'm presuming not) and if not, does the L3 have one? I presume that the L3 switch (internal to subnet) has the IP of say 192.168.5.1, does the aggregation switch ALSO have this IP, or would it be 192.168.5.2?

From there, my understanding is that it say has the external (to the subnet) IP of say 192.168.1.8 while the "Campus Switch" has the IP 192.168.1.1. This would mean there is technically a 'subnet' of 192.168.1.0 between the campus switch, and the 5 pairs of aggregation switches, correct? Then the IP network between the campus switch > firewall > router would be the 192.168.0.0 network (choosing IP for respective ports for each)

Off each L2 switch on each floor/building is multiple Wireless Access Points. I want each building to have its own wireless subnet for students (Subnets 55,56,57,58,59) - separate to the main LAN, but all connect to the internet via web server in the DMZ. If in, say, building 1, all wireless points are subnet 192.168.55.0, does that mean that even if the wireless access point is connected to the same switch as a client on subnet 192.168.5.0, that traffic from the .5 subnet will be ignored, and the data packet received via the WAP will go straight through the network to the L3 switches, then to the campus switch, then the DMZ, then the web server and hence connect to the internet?

Appreciate any help and replies!

Cheers,
Bob