0

I was told one way of isolating wired LAN connected devices from W-LAN connected ones, could be to change their subnets in network adapter settings. By isolating I mean hindering the Wireless connected devices to "see" the LAN ones, or vice versa.

I know where to change the subnets. But my questions are:

  1. Isn't it true that most connections/devices require DHCP automatic IP addresses to function properly? Setting manually a subnet would conflict with that or not?

  2. What implications on the internet connection/security does in general has setting manual subnets instead of letting DHCP handle it automatically?

  3. I don't know what numbers to enter in the subnet mask. How can I know what IPs/parameters should be chosen for IP/subnet? How to do it for ipv4 and how for ipv6? (is it necessary to do it for both ipv4 ipv6?)

  4. Can the DNS parameters stay on automatic mode?

Thanks

1

I would highly suggest for you to look into Subnetting as a network concept. Unfortunately it's hard for me to define and describe the entire concept in a small answer box.

But in summary, subnets break up and delineate Layer 3 networks. This is often notated as such: 192.168.0.0 /24 (or 255.255.255.0 It's just another representation of the bits). In this example 192.168.0.0 is the network address, /24 represents the subnet (that's 24 bits from the left 11111111.11111111.11111111.00000000), and the network address is 192.168.0.255.

Since there is a total of 32 bits in an IPv4 address, we know that we have 8 bits to play with (32-24 = 8). So the total address pool is (2^8)-2, or 254 total addresses.

A different subnet, would just involve setting the network address to something outside of the current subnet. Such as 192.168.5.0 /24.

An easier way to show this is in the binary format. In this example, every bit not "masked" by the positive subnet bits are within the subnet.

‭11000000‬.‭10101000‬.0000‭0101‬.00000000     192.168.5.0
11111111.11111111.11111111.00000000     /24 (255.255.255.0)

‭11000000‬.‭10101000‬.0000‭0101‬.01000100     Same subnet
‭11000000‬.‭10101000‬.0000‭0101‬.00101010     Still the same
‭11000000‬.‭10101000‬.00001111‬.00001000     Different Subnet! (192.168.15.x /24 ≠ 192.168.5.x /24)

We can change subnets just by chaning the masked bits:

‭11000000‬.‭10101000‬.1010‭1101‬.00000000     192.168.173.0
11111111.11111111.11111111.00000000     /24 (255.255.255.0)

We can create smaller or larger subnets by changing the size of the subnet mask:

‭11000000‬.‭10101000‬.1010‭1101‬.00000000     192.168.173.0
11111111.11111111.11111111.11111000     /29 (255.255.255.248)

In general, we create subnets on based on the classes A,B,C

11111111.11111111.11111111.00000000    C 
11111111.11111111.00000000.00000000    B
11111111.00000000.00000000.00000000    A

But this is just convention, and isn't required by any means. But you might hear or read it sometimes.

Network engineers don't memorize every possible combination, but they do understand the logical progression needed to generate what's called a subnet table. I had to create this by hand and in my head on multiple occasions throughout my career. But for everyone else you can just Google Subnet Table and find one that will work for you

Another important network concept you should know is that switches couldn't care less about IP addresses. They only switch traffic based on layer 2 address (MAC Addresses). While IP addresses are on layer 3 and require Routers to route traffic. In general, a LAN will only contain 1 subnet, and you can't have multiple subnets on 1 LAN. However, with managed switches and an Inter-vlan routing capable router you can create VLANs to setup multiple subnets on a single LAN.

Applying this basic knowledge we can start working on your questions:

  1. DHCP is just the protocol that automatically configures network clients. As long as the settings are valid, you can disable DHCP on the DHCP server (most commonly your router) and configure all clients manually if so chose. However, it's important that every client on the network be configured correctly. Typically, every client on a LAN (that's everything on a Layer 2 network typically), must be on the same subnet. In an enterprise setting this is not always the case, as you can implement VLANs and have each VLAN be in a different subnet. But this requires managed switches and very specific planning and implementation.
  2. There would be slightly increased security on the LAN side as each subnet would be in it's own Broadcast domain. However this is really just security through obscurity, and doesn't really do much for real security, but does make management easier some respects. As for setting it manually, it's really just a big pain the ass and huge waste of time. You'd be much better off getting a set of managed switches and creating some VLANs with different subnets and implement DHCP on the VLANS. Lookup "Router on a stick" on Google for information about this network topology.
  3. Subnet's are basically calculated by doing binary math on the binary representation of IP addresses. For the uninitiated, there's tables that provide easy answers. Googling Subnet table is a good help. As for IPv6, it works in a very similar manner but is harder to annoate and explain as it uses hexadecimals. In generally however your subnets should match between IPv6 and IPv4. But since IPv6 has a practically infinite number of addresses I've definitely seen instances where we just hit the "Stateless" (Auto) button on the DHCP server and call it a day.
  4. The DNS setting is just the setting that provides the server addresses for translating domain names to IP addresses. Unless manually changed otherwise, this is usually passed down from your ISP, to your router, to your clients by DHCP. You can set it manually to something like 1.1.1.1 (Cloudflare) or 8.8.8.8 (Google) if you want to, otherwise it will be set via DHCP.

For your implementation of segregated WLAN traffic from LAN traffic by placing them on a different subnet could work, but it would require the Router being capable to routing the traffic for that. Most consumer grade routers are not capable of routing traffic except between the WAN and LAN. However, if your router happens to let you set the DHCP server separately for Wireless traffic, than by all means set it to a different subnet. However, typically you'd have to move up into entry level Enterprise grade systems to gain that functionality.

| improve this answer | |
  • can you please clarify quickly @freebird: (1) "to add 2 routers onto your network" - what is such a device called what do I need to search for at Amazon? -- (2) "Place them "level" with each other" - How do I do this exactly? -- (3) "both WAN ports connect to your modem" -- what is my "modem"? My main WLAN Router I got from my ISP? I do not understand your words "modem" and "router" . I have the WLAN Router from the ISP: What is the second device called that I need to connect with that main router to get a second totally isolated WLan network? Thanks – threeeMiaNichole May 23 at 12:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.