Net Mask Question

[mcp6453]

Accepting the facts as they are, I have two computers on the same
router. The subnet mask is set to 255.255.255.0. One computer is
192.168.1.49. The other one is 192.168.2.49. The first computer needs to
be able to access the second one through the LAN.

I changed the subnet mask on the router to 255.255.0.0. The other
computer was not accessible. Then I changed the subnet mask on the other
computer to 255.255.0.0. The computer was accessible.

Disregarding the reason for keeping the other computer on a different
subnet, what is the disadvantage, if any, of keeping the subnet at
255.255.0.0? Does the router have to be change as well, or does just the
second computer have to be changed? Obviously I don't clearly understand
subnet masks, so an explanation would be very much appreciated.

 

[Bob Willard]

Accepting the facts as they are, I have two computers on the same
router. The subnet mask is set to 255.255.255.0. One computer is
192.168.1.49. The other one is 192.168.2.49. The first computer needs to
be able to access the second one through the LAN.

I changed the subnet mask on the router to 255.255.0.0. The other
computer was not accessible. Then I changed the subnet mask on the other
computer to 255.255.0.0. The computer was accessible.

Disregarding the reason for keeping the other computer on a different
subnet, what is the disadvantage, if any, of keeping the subnet at
255.255.0.0? Does the router have to be change as well, or does just the
second computer have to be changed? Obviously I don't clearly understand
subnet masks, so an explanation would be very much appreciated.

The mask is a filter for IPAs. Incoming packets from IPAs are filtered out
(ignored) if the source IPA does not match the filter. With a mask of
255.255.255.0 (very common), the NIC with IPA=192.168.2.49 will accept all
packets from IPA=192.168.2.x (for all values of x) and will ignore all
packets from IPA=a.b.c.x unless a=192 AND b=168 and c=2.

 

[mcp6453]

The mask is a filter for IPAs. Incoming packets from IPAs are filtered out
(ignored) if the source IPA does not match the filter. With a mask of
255.255.255.0 (very common), the NIC with IPA=192.168.2.49 will accept all
packets from IPA=192.168.2.x (for all values of x) and will ignore all
packets from IPA=a.b.c.x unless a=192 AND b=168 and c=2.

What is the disadvantage, if any, of keeping the subnet at 255.255.0.0?

Does the subnet mask of the router have to be changed as well, or does
just the second computer have to be changed?

 

[Chuck [MVP]]

Accepting the facts as they are, I have two computers on the same
router. The subnet mask is set to 255.255.255.0. One computer is
192.168.1.49. The other one is 192.168.2.49. The first computer needs to
be able to access the second one through the LAN.

I changed the subnet mask on the router to 255.255.0.0. The other
computer was not accessible. Then I changed the subnet mask on the other
computer to 255.255.0.0. The computer was accessible.

Disregarding the reason for keeping the other computer on a different
subnet, what is the disadvantage, if any, of keeping the subnet at
255.255.0.0? Does the router have to be change as well, or does just the
second computer have to be changed? Obviously I don't clearly understand
subnet masks, so an explanation would be very much appreciated.

The subnet mask allows you (the network owner) to adjust the size of your
network, and to segment it into smaller or larger groups, at your convenience.

As long as all devices (computers, and routers) use the same subnet mask, and
have appropriate addresses, using a 255.255.0.0 subnet mask will be perfectly
fine. If you had lots of computers, and wanted to break them up into smaller
groups, you'd use 255.255.255.0, or similar, and assign the addresses
accordingly.

The more different subnets you have, the more routers you need to connect the
subnets. If you have just a few computers on one subnet, you can make the
subnet as small or as large as you wish. Within limits.
<http://www.faqs.org/rfcs/rfc1918.html>
http://www.faqs.org/rfcs/rfc1918.html

But if you want to have one computer with a 192.168.1.x address, and another
192.168.2.x, and you can religiously use a subnet mask of 255.255.0.0, you
should be OK.

--
Cheers,
Chuck, MS-MVP 2005-2007 [Windows - Networking]
http://nitecruzr.blogspot.com/
Paranoia is not a problem, when it's a normal response from experience.
My email is AT DOT
actual address pchuck mvps org.

 

[John B]

--
Sent via OE by John, from MERCURY

What is the disadvantage, if any, of keeping the subnet at 255.255.0.0?

Does the subnet mask of the router have to be changed as well, or does
just the second computer have to be changed?

There is no physical disadvantage. The thing will work. But there is a
"behavioral" disadvantage in that you are deviating from the standard. You
are not taking the well-beaten path. As such, it becomes harder for others
to help you, and to anticipate your presence. To draw a distant analogy,
suppose you want to walk around the block. You can walk on the sidewalk
like most people, or you can walk in the roadway. Automotive drivers might
not expect you to be in the roadway, and you might have an accident.
Similarly, if you want to ride a bicycle, you are advised to ride according
to the "rules," on the right edge of the roadway. If you ride on the
sidewalk, you might get popped by some guy backing his car out of his
garage, because he didn't expect you to be riding a bicycle on the sidewalk.

So when, given your irregular scheme, you hit "reset to defaults" or buy new
equipment, you will have to remember to manually re-enter your non-standard
subnet mask.... It must be the same everywhere on a common ip network.

I recommend you revert to the 255.255.255.0 standard for a Class C network,
which is what your 192.168.x.x private addressing scheme is consistent with.
255.255.0.0 is consistent with a Class B network address. If you revert to
255.255.255.0, then you'll have to put all devices on a common Class C
network for communication to be successful. 192.168.1.x and 192.168.2.x are
DIFFERENT network addresses. That is the primary reason for lack of
communication.

To get devices on different networks to communicate, you must employ a
ROUTING scheme. What you did was change the subnet mask to convert
192.168.1.x and 192.168.2.x to a common network. Another analogy: You had
an automobile with one out of four tires with a smaller radius than the
other three. So you changed one of the three larger tires to be the same as
the one smaller tire. You'll still get to the grocery store, but you'll
look funny getting there. It's not "elegant," to say the least.

Class B, and, furthermore, Class A network addresses are commensurate with
huge, and colossal, networks, respectively. Class C already affords 253
different host addresses to your network. Do you have more than 253 devices
in your network? I doubt that.

 

[mcp6453]

The subnet mask allows you (the network owner) to adjust the size of your
network, and to segment it into smaller or larger groups, at your convenience.

As long as all devices (computers, and routers) use the same subnet mask, and
have appropriate addresses, using a 255.255.0.0 subnet mask will be perfectly
fine. If you had lots of computers, and wanted to break them up into smaller
groups, you'd use 255.255.255.0, or similar, and assign the addresses
accordingly.

The more different subnets you have, the more routers you need to connect the
subnets. If you have just a few computers on one subnet, you can make the
subnet as small or as large as you wish. Within limits.
<http://www.faqs.org/rfcs/rfc1918.html>
http://www.faqs.org/rfcs/rfc1918.html

But if you want to have one computer with a 192.168.1.x address, and another
192.168.2.x, and you can religiously use a subnet mask of 255.255.0.0, you
should be OK.

Chuck, thanks for your response. If I stick with 255.255.0.0, do I have
to change the mask on the router and the computer, only the computer,
only the router? As I posted earlier, I change it on the router, and the
first computer could not reach the second one. When I changed the mask
on the second computer, the first one could reach it.

The question is somewhat pedagogical at this point. The reason I don't
move the second computer into the same subnet is that it is a friend's
computer that I often have to install and service on my network. The
machine has a static IP address that I have to configure for my LAN.
Sometimes I forget to put the IP address back, and he has trouble with RDP.

 

[Bob Willard]

What is the disadvantage, if any, of keeping the subnet at 255.255.0.0?

Answered by Chuck & John B

Does the subnet mask of the router have to be changed as well, or does
just the second computer have to be changed?

With the router at the (standard) IPA/mask of 192.168.1.0/255.255.255.0,
only the PC with IPA=192.168.1.49 will be able to communicate with the
router; the PC with IPA=192.168.2.49 will not able to control the router.

 

[Chuck [MVP]]

Chuck, thanks for your response. If I stick with 255.255.0.0, do I have
to change the mask on the router and the computer, only the computer,
only the router? As I posted earlier, I change it on the router, and the
first computer could not reach the second one. When I changed the mask
on the second computer, the first one could reach it.

The question is somewhat pedagogical at this point. The reason I don't
move the second computer into the same subnet is that it is a friend's
computer that I often have to install and service on my network. The
machine has a static IP address that I have to configure for my LAN.
Sometimes I forget to put the IP address back, and he has trouble with RDP.

You have to change it on ALL devices - computers and routers - identically, if
you want each to communicate with the others.

--
Cheers,
Chuck, MS-MVP 2005-2007 [Windows - Networking]
http://nitecruzr.blogspot.com/
Paranoia is not a problem, when it's a normal response from experience.
My email is AT DOT
actual address pchuck mvps org.