Subnets?

[nonnormman]

If I subnet my class c network with 255.255.255.240 then I should have 16 subnets. If I do this from wtih my router, how can I place certain machines on my network on a different subnet? When I change my subnet mask to 255.255.255.240, I get the first subnet of the range 1-14. How could I take advantage of ranges 17-32 and so on?

 

[bruceb]

Downlaod this free program .. IPSUBNETCALC ... it can be used to help
determine what is what on subnets

http://www.wildpackets.com/products/free_utilities/ipsubnetcalc/overview

 

[ScottMac]

You have to have multiple interfaces, one for each subnet.

The interfaces can be real, or they can be "secondary" (usually a silly thing for anything except migration and a possible security risk).

Good Luck

Scott

 

[nonnormman]

You mean if I add a dhcp server, I can have it assign ip addresses in the 17-32 range on the 255.255.255.240 subnet mask?

 

[ScottMac]

Some DHCP servers can serve up multiple scopes (scope ~subnet range and parameters specific to that segment).

In order to use (for example) four subnet blocks you need for interfaces on which to apply them.

For example, if you had a Cisco router with four ethernet interfaces, you could use one of the subnet blocks on each of the Ethernet interfaces. IN teh case of CIsco routers, you can add additionaol blocks of addresses to an interface with a "SECONDARY" keyword (for example, you could have a 10.1.1.0/24 address as the primary block, and a 172.16.0.0/16 address as a "SECONDARY" block.

This is usually used when you are migrating from one address scheme to another, so all of the clients can still function, even though some are on one address block that the rest are on another. In any other practical context, it's a silly thing because it doesn't really accomplish anything.

Otherwise, the subnetting example you've posted would allow you to break one large address block into several, which would usually be used to permit several locations to use a portion of the original block / address space ... or several LAN segments .... but each segment needs its own interface on the router.

If you have a standard SOHO router with a WAN interface (to a BB or DSL modem) and an Ethernet interface (usually ~four or eight switch ports on the same routed interface), then subnetting would do you no good .... you only have one interface for which you can select the address space.

You can subnet and reduce the potential count of hosts on the network; the .240 mask gives you addresses for fourteen hosts ... that would be all you can use ... if you used an address outside of that block, it would be an invalid address and the client couldn't communicate.

What are you trying to do, and what equipment are you trying to do it with?

FWIW

Scott

 

[nonnormman]

Originally posted by: ScottMac
Some DHCP servers can serve up multiple scopes (scope ~subnet range and parameters specific to that segment).

In order to use (for example) four subnet blocks you need for interfaces on which to apply them.

For example, if you had a Cisco router with four ethernet interfaces, you could use one of the subnet blocks on each of the Ethernet interfaces. IN teh case of CIsco routers, you can add additionaol blocks of addresses to an interface with a "SECONDARY" keyword (for example, you could have a 10.1.1.0/24 address as the primary block, and a 172.16.0.0/16 address as a "SECONDARY" block.

This is usually used when you are migrating from one address scheme to another, so all of the clients can still function, even though some are on one address block that the rest are on another. In any other practical context, it's a silly thing because it doesn't really accomplish anything.

Otherwise, the subnetting example you've posted would allow you to break one large address block into several, which would usually be used to permit several locations to use a portion of the original block / address space ... or several LAN segments .... but each segment needs its own interface on the router.

If you have a standard SOHO router with a WAN interface (to a BB or DSL modem) and an Ethernet interface (usually ~four or eight switch ports on the same routed interface), then subnetting would do you no good .... you only have one interface for which you can select the address space.

You can subnet and reduce the potential count of hosts on the network; the .240 mask gives you addresses for fourteen hosts ... that would be all you can use ... if you used an address outside of that block, it would be an invalid address and the client couldn't communicate.

What are you trying to do, and what equipment are you trying to do it with?

FWIW

Scott

I want my computer to be on a different subnet then the other computers.

 

[Fardringle]

Then you will need to have a DHCP server capable of serving separate subnets.

The real question is, why do you want to do this? Putting it on a different IP address range will not prevent the other computers from communicating with your PC, or your PC from communicating from them as long as they are all on the same physical network. If that is what you want, then you need a VLAN (virtual LAN), or a separate router/firewall that will isolate your computer from the rest of the network, or you need to physically put it on a different network. If you just want to keep the other computers on the network from communicating with your PC, you can also just install a good software firewall on your PC and block all incoming traffic.

 

[Brazen]

Originally posted by: nonnormman

Originally posted by: ScottMac
Some DHCP servers can serve up multiple scopes (scope ~subnet range and parameters specific to that segment).

In order to use (for example) four subnet blocks you need for interfaces on which to apply them.

For example, if you had a Cisco router with four ethernet interfaces, you could use one of the subnet blocks on each of the Ethernet interfaces. IN teh case of CIsco routers, you can add additionaol blocks of addresses to an interface with a "SECONDARY" keyword (for example, you could have a 10.1.1.0/24 address as the primary block, and a 172.16.0.0/16 address as a "SECONDARY" block.

This is usually used when you are migrating from one address scheme to another, so all of the clients can still function, even though some are on one address block that the rest are on another. In any other practical context, it's a silly thing because it doesn't really accomplish anything.

Otherwise, the subnetting example you've posted would allow you to break one large address block into several, which would usually be used to permit several locations to use a portion of the original block / address space ... or several LAN segments .... but each segment needs its own interface on the router.

If you have a standard SOHO router with a WAN interface (to a BB or DSL modem) and an Ethernet interface (usually ~four or eight switch ports on the same routed interface), then subnetting would do you no good .... you only have one interface for which you can select the address space.

You can subnet and reduce the potential count of hosts on the network; the .240 mask gives you addresses for fourteen hosts ... that would be all you can use ... if you used an address outside of that block, it would be an invalid address and the client couldn't communicate.

What are you trying to do, and what equipment are you trying to do it with?

FWIW

Scott

I want my computer to be on a different subnet then the other computers.

Why? Do you want to put it behind a router? Do you want your computer to still be able to talk to the other computers?

Your router will need a second IP address on the same subnet as your computer in order to access the router. You don't need to mess with your DHCP server (you would have to be behind a router with DHCP relay enabled, anyway, for that to work). Just assign a static IP address to your computer and give it the subnet mask you want.

To really answer your question though, we need more info. Like I asked above, are you putting your computer behind it's own router? Are you wanting to completely block access to and from your computer, with only access to the net and not other computers? Physically, what kind of setup are we looking at here, and what is the purpose of putting your computer on a different subnet?

I would like to point out that Fardringle also gave some false information. Putting your computer on a different subnet WILL completely remove any ability to communicate with other computers, to or from, even if on the same physical network, unless a virtual ip and route is set up on your gateway.

 

[kevnich2]

Yes, Fardringle is very wrong with that statement. Whether or not your on the same physical network, if you have any computers on a separate subnet, communication is non-functional unless you have a router to communicate with that subnet. I really cannot think of any reasons to do this?

 

[ScottMac]

Putting your computer on another subnet, just for the sake of doing it?

If you have a (many / most / all?) SOHO router, you can't.

If you have some Cisco router(s) (or Synoptics/Bay/Nortel, 3COM, Retix, Juniper ...) and you want to "practice subnets" is a general waste of time (not much practice .... I guess to verify your config would be somewhat valid).

If it's for some flavor of bragging rights ... well, in the words of my grandfather: " There ain't much braggin' rights in being stupid"

Tell us what you have to work with, hardware-wise and we can take it from there.

You don't need another DHCP server ... you don't need any DHCP server at all, anywhere, ever, at any time. They're convenient, but not mandatory.

Good Luck

Scott

 

[nonnormman]

I just wanted to get some idea on how business networks are set up.

 

[Brazen]

Originally posted by: nonnormman
I just wanted to get some idea on how business networks are set up.

It depends on the situation. We use 4 different subnets, all with different subnet masks (that only affects the number of available ips within a subnet). 1 is our main location, 2 are for branch locations, and 1 is for a department at our main location that is firewalled from the rest of the organization (for security reasons). Since these all have to be behind routers, we have to use different subnets for each of them.

Would you like more explanation on what subnets and subnet masks are? and how you use your router to access them? I can try to write up a simple explanation, but takes a lot of explaining, so I will do it, but only if you want me to .

 

[ForumMaster]

do a google for a program called Packet Tracer 3.2. it let's you simulate networks an i use it in my cisco class. it's a great program and let's you test your config before setting it up for real.

 

[nonnormman]

Originally posted by: Brazen

Originally posted by: nonnormman
I just wanted to get some idea on how business networks are set up.

It depends on the situation. We use 4 different subnets, all with different subnet masks (that only affects the number of available ips within a subnet). 1 is our main location, 2 are for branch locations, and 1 is for a department at our main location that is firewalled from the rest of the organization (for security reasons). Since these all have to be behind routers, we have to use different subnets for each of them.

Would you like more explanation on what subnets and subnet masks are? and how you use your router to access them? I can try to write up a simple explanation, but takes a lot of explaining, so I will do it, but only if you want me to .

I understand subnets and masks, but I was just curious to learn more about how bigger networks subnet their networks. Thank you all for the help.

Originally posted by: ForumMaster
do a google for a program called Packet Tracer 3.2. it let's you simulate networks an i use it in my cisco class. it's a great program and let's you test your config before setting it up for real.

Thanks.

 

[spidey07]

Larger networks normally have a hierarchial design.

There is a "core" that is interconnected by high speed links. These are large and powerful routers with very high speed connections. Their job is to move packets as fast as possible to the next layer...distribution.

distribution layer normally has two or more connections to the core for redundancy. It's role is to centralize all the routers at the access or edge of the network. Your cable modem or DSL is connected to a network off of this layer - the edge/access.

How is all this subneted? Normally there is some kind of hiearchy in the address space design to keep routing tables small and mangable If we're talking about a LAN it follows the same design somewhat. In the LAN arena its customary to use a single class C network per floor/access layer. Use of a class C keeps broadcasts contained and just makes sense.

All of the routers connected to each other use a mask of 255.255.255.252 or 255.255.255.248 for these links. Since there are really only two ip addresses used - router a interface and router b interface.

Same model can be used in large data centers as well. You have a core and then some multiple access layers for servers. The different server access layers then communicate to each other through the core.

So after all the hullabullu, larger networks normally use a class C at the edge and then use /29 or /30s to interconnect the routers.

If you want to learn more google "large scale network design"

 

[nweaver]

DHCP can be easily pushed to different subnets using iphelper, and defining the gateway on an MS DHCP server. I know, I have one server, (single IP) serving 8ish subnets/scopes.


can I just say, I like Scott's Grandpa!