rslinx and ip addressing

The first and last addresses in a TCP/IP subnet are reserved (not for use for your devices). The former identifies the "network", while the latter is used to send broadcasts to everyone. Suppose you're using 192.168.1.x addresses. .0 represents the network and .255 is the broadcast address. This assumes a 255.255.255.0 subnet mask. Change that and the network/broadcast addresses change as well.

In networking, a "node" is a device that connects to the network. For example, a PC, printer, server, PLC, router, etc.

MannyM said:

Please define network and broadcast with respect to either bit assignment or eitherwise. Node from a design point of view, should remain the gate to open. Beyond that, it is quite another story.

Please rectify me if I am wrong on this subject.

Click to expand...

The subnet mask determines which bits determine the network versus the host. I've heard stories of network administrators creating "job security" with screwy schemes like that. Computers would work just fine, but the numbers of your IP addresses in the same network wouldn't be sequential. Basically, it's confusing with no benefit.

Also, if you're on a larger network, it would create inefficiency if not problems with routing. Routers have a feature called "route summarization" where they remember a "whole network/summarized range" as a single entry in their route table. Bad addressing forces them to remember each node as a separate entry.

jonnies said:

I have a purely theoretical question about this topic. It seems like subnet masks, in binary, are always split into 1's on the left and 0's on the right. What would happen if a subnet mask was something like 255.255.31.0, or 11111111.11111111.00001111.00000000? Would the network ID be the first two values, and the last half of the third value? Or is it only read until the first 0 is found?

Click to expand...

Nathan

Will skipping lage blocks of addresses cause this problem also?

Example 10 nodes on the network 192.168.1.1,192.168.1.20,192.168.1.30,and so on.

Will this cause lag in the routing?

Subpar routing table?

No - that's fine. As long as they're on the same network the routers will just use the summary. For example, if your network is 192.168.1.0/24 (meaning 1-255), then any address that begins with 192.168.1.x will be properly routed.

As an exaggerated point, consider if some clown set a network so that 1.10.10.15, 23.17.10.15, and 189.3.2.254 are on the same network, but 1.10.10.35, 23.17.10.16, and 189.3.3.254 are on a different network. Computers wouldn't have a problem with this - it confuses the heck out of people because it makes no sense. Suppose we have a large internetwork with routers in between. Each router would have a rediculous load because it would have to remember so many individual hosts. Older routing protocols (like RIP) wouldn't even support this nonsense.

plc noob said:

Nathan

Will skipping lage blocks of addresses cause this problem also?

Example 10 nodes on the network 192.168.1.1,192.168.1.20,192.168.1.30,and so on.

Will this cause lag in the routing?

Subpar routing table?

Click to expand...

How would you put 1.10.10.15, 23.17.10.15, and 189.3.2.254 on the same network.

In this example the subnets are all different.Is it the first 2 or 3 octets that makeup the subnet?

Also i have never seen an address setup this way 192.168.1.0/24 (meaning 1-255), Where and why would you enter it like 192.168.1.0/24?

Sorry for all the questions but i am self taught ethernet and learned most of what i know from this site and primarily from your posts nathan.

Your posts are always very helpful and educational.

Thanks

There are many screwy subnet masks that might work. For example a subnet mask of 0.0.2.0 would put those three nodes on a huge super-network. Don't sweat the details on that one.

plc noob said:

How would you put 1.10.10.15, 23.17.10.15, and 189.3.2.254 on the same network.

Click to expand...

The subnet mask tells you which octets (or parts of octets) represent the network versus hosts. The "/" slash notation is shorthand for the number of bits that are leading 1s. Remember an octet, 255 in decimal, is 8 bits.

For example:
(Class C) subnet mask, 255.255.255.0, or /24, means that the first 3 octets identify the network id and the 4th is for hosts. This gives you (roughly) 2^24 networks and 2^8=256 hosts. Any addresses with the same first three octets are on the same network.
(Class B) subnet mask, 255.255.0.0, or /16, means that the first 2 octets identify the network id and the next 2 are for hosts. This gives you (roughly) 2^16 networks and 2^16=65k hosts. Any addresses with the same first two octets are on the same network.

plc noob said:

In this example the subnets are all different.Is it the first 2 or 3 octets that makeup the subnet?

Also i have never seen an address setup this way 192.168.1.0/24 (meaning 1-255), Where and why would you enter it like 192.168.1.0/24?

Sorry for all the questions but i am self taught ethernet and learned most of what i know from this site and primarily from your posts nathan.

Your posts are always very helpful and educational.

Thanks

Click to expand...

I hope this helps instead of confuses. It's a bit hard to wrap your mind around until you realize that you should be thinking in binary. We just convert to decimal to remember the numbers more easily.

The more common application of subnetting is cutting IP address ranges. Suppose you wanted to allocate 30 addresses for something. You'd use a subnet mask of 255.255.255.224 or a "27 bit subnet mask". So, 192.168.1.0/27 gets the range 192.168.1.1-31. The number of addresses came from 2^(32-27). The number of networks came from 2^27. The "224" came from the leading 3 bits of the 4th octet (128+64+32).

Clear as mud?

Thanks Nathan

That does help.I want to know all the details. Any good books on the subject that you would recommend?

Just do a Google Search on "IP subnetting" the first link that came up is a pretty good introduction.

http://www.ralphb.net/IPSubnet/

This is good too:
http://articles.techrepublic.com.com/5100-10878_11-6089187.html

Simple explanation :-

A Subnet Mask can be any bit pattern you want on a private network.

For devices to successfully talk to each other - 2 rules, Same Network, Different Node

1. Where the SubNet Mask has a "1" bit (this is the "Network ID"): same bit pattern in the device IP addresses.

2. Where the SubNet Mask has a "0" bit (this is the "Host ID"): different bit patterns in the device IP addresses.

Read as much as you can, it will add to your understanding.

And FYI, the reason why 192.168.xx.xx seems to crop up so often, is that it is one of a stock of reserved Private IP addresses, and Routers and Gateways will never allow data packets containing those addresses outside of the private network, i.e. onto the Internet