Any harm in using a class A network?

T

Tolem

n00b
Joined
May 27, 2010
Messages
33
Every router (soho) I have used always defaults to a class C or 192.168.x.x network. I like to use 10.0.0.0 just because its less numbers (faster to type, easier to remember IP's, just simpler)

I was wondering if there was any harm or downside to this?? What about when dealing with enterprise level networks??
 
you can use the 10.0.0.x ip range in a class C network fyi.

On a home network? Its just pointless, won't really cause issues one way or the other, on a large network, BuHAHAHAHAHHAHHAH, unless you have a specific reason, you'll have management headaches up the ying yang
 
why would there be management headaches up the ying yang to use 10.0.0.* on a larger network?

it is allowed same as 192. ?
 
If you use vpn you have to be sure it is not getting used in your corporate. if you use multiple routers (not a good idea) you have to assign each different ip ranges and subnets. Aside from that, no issues.
 
nah, your good to go.

we use 10.x.x.x for my corporate network here.

no issues. 10.1.X.X is used for the different vlan traffic, etc. (separates voice/data traffic)

also we use 10.2.X.X for a different company that we manage as well, all get tied into our ASA 5510

holla
 
I would limit your subnet size to a /24 at the largest.

PLAN OUT YOUR IP SPACE. I can't stress this enough. Don't just go in willy nilly, you ( or your replacement ) will regret it later.
 
I'm using 10.100.0.0 at home internally. I haven't seen a corporate network yet that was large enough to where I'd end up duplicating any internal IP address space while on VPN.

I suppose someone could go crazy and give every remote site they had a full 24 bit network mask range even if they only have 5 or so fewer machines at each remote site, but I haven't come across it...
 
Just curious: How come the most used mask is /24?? Its for 254 host addresses right? What kind of bad things happen if you use larger than 24, such as /20 or .. comedy option /8??

But I see /24 the most, what about smaller masks like /27 and the smallest is /30 right or 2 hosts?

:confused:
 
Class A, B, or C doesn't really matter since nobody has used classfull routing for years. It's all CIDR/VLSM, so you can have a /24 in a Class A or B just fine. FYI /24 is native for Class C anyway.

Look up RFC 1918 for more info on private networks, and there's plenty of resources online for IPv4 subnetting.


Tolem said:
Just curious: How come the most used mask is /24?? Its for 254 host addresses right? What kind of bad things happen if you use larger than 24, such as /20 or .. comedy option /8??

But I see /24 the most, what about smaller masks like /27 and the smallest is /30 right or 2 hosts?

:confused:
Click to expand...

Just more available hosts per subnet, which you'll have fewer of. Generally speaking you don't want more than 500 hosts in a single protocol broadcast domain due to overhead, so you'd want to subnet. If you're running multiple protocols then try to limit the broadcast domain to 250 hosts. If this is only for a home network then it doesn't actually matter what you subnet to.
 
BobSutan said:
Class A, B, or C doesn't really matter since nobody has used classfull routing for years. It's all CIDR/VLSM, so you can have a /24 in a Class A or B just fine. FYI /24 is native for Class C anyway.

Look up RFC 1918 for more info on private networks, and there's plenty of resources online for IPv4 subnetting.




Just more available hosts per subnet, which you'll have fewer of. Generally speaking you don't want more than 500 hosts in a single protocol broadcast domain due to overhead, so you'd want to subnet. If you're running multiple protocols then try to limit the broadcast domain to 250 hosts. If this is only for a home network then it doesn't actually matter what you subnet to.
Click to expand...
Uh, what he said.

Again, I have to stress; if you are going to vlan/subnet, plan this shit out in advance. Setup a vlan for servers, a vlan for network management, a DMZ vlan, an extranet vlan, ect...Even if you don't think you'll use them, plan it out on paper so should you need to your schema is laid out already.
 
I mapped out my network on visio a long time ago, and keep it updated...

What are the benefits of having everything on a separate vlan vs as I have it now where servers have an IP range, network devices have an IP range, etc.
 
Tolem said:
I mapped out my network on visio a long time ago, and keep it updated...

What are the benefits of having everything on a separate vlan vs as I have it now where servers have an IP range, network devices have an IP range, etc.
Click to expand...

keeping traffic separated man.
 
Tolem said:
I mapped out my network on visio a long time ago, and keep it updated...

What are the benefits of having everything on a separate vlan vs as I have it now where servers have an IP range, network devices have an IP range, etc.
Click to expand...
Traffic management. For instance, you generally don't want your network devices to be initiating connections to your workstations. Same with your servers.

By separating the different kinds of traffic out by subnet, you simplify management.
 
I didn't know they would initiate connections on their own...?

If I do this and I'm at a workstation, then I want to check on a server can I still RDP into it as if they were on the same subnet / vlan or will I have a problem?

When you said a vlan for network management, what exactly falls under "network management" a workstation with access to things like RDP and programs to monitor health machines / network???
 
Tolem said:
I didn't know they would initiate connections on their own...?

If I do this and I'm at a workstation, then I want to check on a server can I still RDP into it as if they were on the same subnet / vlan or will I have a problem?

When you said a vlan for network management, what exactly falls under "network management" a workstation with access to things like RDP and programs to monitor health machines / network???
Click to expand...
They shouldn't initiate connections on their own, that's the point. By separating out the traffic, it is easier to implement traffic restrictions should the need ever arise.

Not to mention there's a large part of me that's anal-retentive. I like to see similar workstations within the same IP space ( perhaps a /27 or something ). Even if it's not officially subnetted out, I group my workstations by dept/purpose.

On it's own, subnetting your network doesn't automatically implement traffic restrictions, and you often need a device capable of doing filtering if you want to implement restrictions. Otherwise you'll simply have different IP spaces and a router to pass the traffic. But it's still worth planning out your vlans in advance if only to have a plan.
 
Ok so following your advice I begain to use the vlsm calc and visio to plan out a subnet diagram which will keep everything in its own little corner of the network :D

But something I noticed was, the largest subnet (/25 in my case) was assigned (by the calculator) the first block of IP's such as 10.0.0.1 - 10.0.0.126 when in reality I have the servers as 10.0.0.1-10.0.0.10 since its so easy to remember. The block I was going to put them on (a /27) ended up being after 10.0.0.230 something since it was the smallest block.

Is there a way around that or is that just how subnetting works?
 
I would stick with /24s in practice, it's what everyone will be looking for. I only brought up the /27s in reference to how I place workstations by purpose/dept. I don't actually subnet down to /27, I just assign them addresses as if it were subnetted by /27 ( or /28, or whatever fits ).

Again, that's my own anal-retentive nature. The important thing is to plan out in advance your larger scopes ( workstations, servers, network devices, dmz, ect.. ). Everything else is extra.
 
It really depends on who will be looking after the network, if it's just you, and other people that actually know what VLSM is, then use whatever mask is appropriate (/30's on point to points for example), the problem comes when you get someone in that's only ever seen a /24, and just assumes every subnet mask = 255.255.255.0.

the reason why you don't use 10.0.0.0 /8 is because you would have a huge broadcast domain, do you really expect to have (2^24)-2 hosts?

Also, to whomever it was that mentioned they wanted to assign their hosts particular IP's that would normally be the first few in a /24, (10.0.0.1-10 or whatever) in their VLSM network, it really depends on the network you're actually using.
totem said:
But something I noticed was, the largest subnet (/25 in my case) was assigned (by the calculator) the first block of IP's such as 10.0.0.1 - 10.0.0.126 when in reality I have the servers as 10.0.0.1-10.0.0.10 since its so easy to remember. The block I was going to put them on (a /27) ended up being after 10.0.0.230 something since it was the smallest block.

Is there a way around that or is that just how subnetting works?
Click to expand...

Say you have two /25's

Network 10.0.0.0
subnet mask 255.255.255.128
valid host range 10.0.0.1 to 10.0.0.126
broadcast 10.0.0.127

network 10.0.0.128
mask 255.255.255.128
valid hosts 10.0.0.129 to 10.0.0.254
broadcast 10.0.0.255

but now you really want your servers to be in a /27, and you want them to use the ip's 10.0.0.3 and 10.0.0.4

a /27 gives you 32 addresses, 30 usable ( (2^5)-2 )
you would need to carve out a subnet from the 10.0.0.0/25 (which you could no longer use as a full /25, you'd need to carve additional subnets out)
so you would have:

network 10.0.0.0
mask 255.25.255.224
valid hosts 10.0.0.1 to 10.0.0.30
broadcast 10.0.0.31

and your next usable network would be 10.0.0.32/27 and so on.

disclaimer - I typed that out pretty quick, might have made mistakes, look it over before implementing :p
 
Thanks for the refresher in VLSM subnetting, been a couple of weeks since I looked at it. I just want to skip over the A+ book and go back into my CCNA book, much more interesting.
 
You must log in or register to reply here.
Back
Top