Tricky Networking Question

- "Steve administers a network with following addressing scheme
192.168.0.0/25 Default gateway 192.168.0.1"

- "Only change steve makes is to turn on a printer which has an IP address of 192.168.0.121/24 default gateway 192.168.0.1."

Just for clarification, the printer (192.168.0.121/24) is using a subnet mask that is different from Steve's addressing scheme (192.168.0.0/25). Is this correct or was there a typo? I just wanted to double check first!

Also do you know if the two networks are using VLANs?

Sorry I just noticed that there was a response to my question.

I agree with both of you as well. The amount of information that is given, isn't enough to provide a legitimate explanation.

Magician27, the reasons that you listed for why a local device might not reply to a Ping are correct. A Local device that doesn't respond a Ping is usually the result of a firewall restriction and/or the router (or the device itself), is configured not to respond to ICMP echo requests.

As bill001g said, the configurations would have to be pretty over the top. I also agree with you both regarding the issue of the two networks being on separate Broadcast Domains. This is why is asked if the networks were using VLANs. This question is implying something that is fundamentally incorrect based on the information given (unless I'm completely missing something).

Technically speaking, it is possible to configure different broadcast domains (e.g. different subnets,IP address ranges) with the same VLAN ID, thus making that VLAN a broadcast domain of its own.

However, it's usually not good practice to configure multiple IP subnets to share a single VLAN. Most will recommend against doing this and it's rarely used bc its not a practical design. Just in case anyone who reads this thread wants to know how I'll try to explain. Please feel free to clarify or point out anything that's incorrect:

A VLAN is basically a Broadcast Domain that is logically segmented based on an arbitrary number of grouped ports, that together make up an autonomous workgroup. Most people recommend having a "one-to-one" relationship between VLAN and Subnet; in other words a VLANs should only include devices on the same subnet.

It is still possible to have multiple subnets on the same VLAN, however there are some issues with this type of design. Simply put VLANs are layer 2 and IP subnets are a layer 3 function. Different subnets each have their own broadcast address, and broadcasts will not outside of that network (i.e. 192.168.2.0/24 >> 192.168.2.255). Likewise a different subnet will have it's own broadcast address and it's broadcasts won't go outside of their network (i.e. 192.168.100.0/24 >> 192.168.100.255).

In this example it's obvious why they are separate broadcast domains. Broadcasts Domains consist of all devices on a segment that receive a network devices broadcast message. However, if a devices from the two subnets belong to the same VLAN, they will also be apart of the same Broadcast Domain. However, this will only work if the two subnets are properly interconnected.

Since VLANs function at layer 2, and IP subnets are layer 3, these two subnets can only use the same VLAN ID if they are separated by a layer 3 router. The layer 3 boundary allows to different layer 2 broadcast domains to communicate by trunking from each individual subnets own switch, to the router. When two hosts belonging to the same VLAN, but different subnets wish to communicate, messages will be sent to the router to be forwarded as packets. The router provides the layer 2 switches with the layer 3 functionality required to communicate with different IP subnets.

Adding two subnets to the same VLAN also creates the issue of broadcasts possibly reaching the wrong destination. IP is a broadcast technology as well, which means different subnets sharing a VLAN, technically have two broadcast domains.

In my earlier example, a device on the subnet 192.168.100.0/24 would have broadcast domain192.168.100.255 and broadcast domain VLAN10 that it shares with hosts from different subnets. A broadcast intended for VLAN 10 might accidentally arrive at a destination belonging to the 192.168.100.0/24 domain.

In this case Devices on 192.168.100.0/24 that don't belong to VLAN10 might receive the message, which is a security risk. ARP, DHCP, and routing protocols all rely on broadcasts, and would likely present many issues using a set up like this.

Again there is rarely a situation that would justify adding multiple subnets in a single VLAN.

I suspect you will find that there is some assumption that there is no way to guess.

2 devices on the same subnet talk DIRECTLY to each other via mac address, only the switch would be involved. The only way they can not communicate is if there is a issue with one or the other. The router or devices on another network are not in the traffic path.

When you find out what the idiot who wrote this question thinks the answer is you can come back and post it.

Even if you could why would the traffic ever go to the router in the first place. The traffic is totally withing a subnet it goes end machine to end machine via mac address. Nat is all IP based. It is extremely simple process. Machine doing the PING issues a ARP broadcast the other machine either responds or does not. If it does then the ICMP packet is built and set to the mac address. You either are losing the broadcast or the switch is not forwarding the ICMP correctly. None of this function even needs the router to do anything it is pure layer 2 in function so no traffic will pass the router.

It can have no effect it must be on a different switch and interface on the router. Broadcasts can not pass between them this is fundamental to how layer 2 and layer 3 traffic work.

Again they have to assume some strange very non standard cabling or configuration.

I can make up lots of things... I will just say they had turned off spanning tree on the switch and cabled 2 ports together. Maybe they had a VPN configured between the machines. Maybe they were running a bunch of virtual routers on the router and configured it wrong.

There is no way to guess what can cause a problem if you allow the possibility of international misconfiguration or using very non standard configurations.