If running more than one subnet is a good design, rather than one subnet, I could probably use VLANs, now that I have some sort of idea how to. I could assign the VLAN to the port that clientrouter is connected to, and assign it to it's own subnet, with it's own DHCP router. However, the downfall of this, is that if I did that, then I'd need to make sure that multicast was forwarded to the VLAN, and that guestrouter knew about it, and that I did not do anything else to break airprint, now that it's working. If it's a good design, there might also be a way to set it up just in clientrouter, without having anything else affected.

I'll do some quick research.

I found these:

https://www.dd-wrt.com/phpBB2/viewtopic.php?p=537038
https://www.dd-wrt.com/phpBB2/viewtopic.php?p=878670

And I don't know, but this:

https://www.dd-wrt.com/phpBB2/viewtopic.php?p=1009873

...

...May have been what I wanted...

Well, if that works, it would solve the original thing I may have wanted to do, but I don't know about the bigger picture.

While I can't find what the routing setting does, very well, I can now see one theory on how to put my router in router mode and have it work. It looks as if I add a gateway setting to the routing table of the first router.

However, if that's the case, would multicast packets make it through from one subnet to the other? As I said before, I "think" that that was the problem with airprint not working before, according to the forum I was previously on. So if I could have my subnet, and my airprint would work too...

So the next question, is this then, since we are getting closer to a solution. Which is a better design, to have? Is it a better design to have 2 subnets, so that I can have 2 DHCP servers at work, serving separate IPs for clients and basic network hardware? Or is it a better design, since I do not yet have more than 254 clients on the network, to use one subnet, and one DHCP server?

A helpful factor in deciding, is that there is always some IP switching to do, when adding clients/servers properly, as per the design.

According to the design, the MAC never changes unless there is a conflict, but may be recorded at times. The IP, is mostly static, but will sometimes change what that static IP is, as new devices are added or subtracted from the network.

The DNS name will not change, unless there is a new device from that owner of the same type.

---
For example, let's say I get a new laptop to use for completely personal use, and I want to put it on the network.

I will first need to make sure I have permission to add a device. I have already signed an updated AUP, so we are good to go there. I will need to provide my MAC, which I can change if I want more privacy or if there is a conflict. If the MAC changes, I will need to update it.

Then, I can be assigned an IP. I will follow the network standards to decide who gets which IPs, and then set up the corresponding router accordingly, which may shift existing IPs up or down.

Then, I will get a DNS name. Let's say this is a client. Let's also say it's another laptop computer. So my IP will be changed in the place for clients, which is now in mainrouter, but was in clientrouter before. I own more than one device already, including more than one computer. So my DNS name is a number in my special format (which is often a letter), w, if it's a wireless interface, then a dash, then my 3 initials. It will not be required for the network to work, so there will be no dash and then no "rec".

The number will be the number corresponding to the order the computer was obtained in. So if this is my 3rd computer, the DNS name would be "c-des".

These are updated, if the order of the computer changes, such as getting rid of a computer.
---

I could let users request a special alias, to their client machines, but this would be upon request, not given standardly. The alias, would not have to change.

P.S. - By the way, the switching is why there ARE standard aliases for most servers. That's where mainrouter, clientrouter, and guestrouter come from. They are basically treated as a type of server. Those are the shortest names for them, and therefore, the easiest to remember. Those are their hostnames. I do believe, that as of yet, none of them have WINS names, just hostnames. But some of them might have WINS names later, because some of them will also become NASs.

The 3 domain names, are smiley000.net, clients.smiley000.net, and guests.smiley000.net. Now, clients.smiley000.net is the same IP range as smiley000.net, while guests.smiley000.net is a different range.

mainrouter, is in smiley000.net. clientrouter, is in both smiley000.net, and clients.smiley000.net. guestrouter is in both smiley000.net, and guests.smiley000.net.

These are internal names, NOT external names. They won't work on the Internet at large, just my network. On the Internet at large, my only name that can be used right now, is smileynet001.ddns.net.

I could mess around with names on the outside, but that'd cost me money to register the domain names, so instead I stick with free solutions for now.

OK. Here goes a reconfiguration experiment. I will attempt to reconfigure things again, and see if I broke airprint or not. Please be patient. I'm no longer home right now. I'll reconfigure it from here, and when I get home, I'll check it.

All went ok, except for when I added the route to the routing table. It looks like I added it wrong. I'll have to read up on routing tables, I guess. At first, I added

But since that didn't show up in the routing table, I tried to add:

Everything then didn't work. I'll have to fix it when I get home. I'll probably have to reset the router, and then re-install the configuration settings. It'll be awhile until I get home again...

Then, of course, when I get this working, I'll see if I broke airprint or not. But NAT should be disabled.

...and it should be in it's own subnet.

I found these links:

https://www.techrepublic.com/article...outing-tables/
http://library.mobrien.com/Manuals/M...es/routing.htm

So, the reason it's now not working, is because I told it to change the default gateway. So, it's probably functioning, but not from here, and without Internet. But why didn't it change the routing tables?

I could always try from the command line, but that seems silly, if there's a GUI that works...

des_a. I was wrong. I said before I was lost, now I am really, really lost.

Is there any chance you can re-edit all this to some simple question while deleting all the extra information?

I found these:

https://www.dd-wrt.com/help/english/HRouting.asp
https://www.dd-wrt.com/wiki/index.ph..._Static_Routes

This:

https://www.dd-wrt.com/wiki/index.ph..._Static_Routes

...is probably what I should have done...

Oh well!

THE QUESTION RESTATED
=====================

I'll try to add information first, which will hopefully tell you, what you need to know.

What I was trying to ask, in the most simplest form, was how to get clientrouter on a subnet of it's own, without NAT. But then they should be able to communicate with each other, back and forth. Unfortunately, I did not know the terminology to ask that yet. I was gaining the terminology as I troubleshooted. I still don't know the exact terminology, in technical terms, so that description will have to do. I knew it was when the router was in Router mode, that this can happen. But I didn't figure out before, how to make things like Internet work, which I still wanted to work.

My larger purpose, of asking this question, was to create a system where airprint would work, without doing anything like hooking things up differently, or getting new equipment. The underlying problem of that, was that multicasting wasn't working with NAT. I wanted to see if disabling NAT would make things work or not. It may or may not, as my experiments and learning have found. I got airprint to work with everything in the same subnet, the question now, is the way I wanted to try it first, to begin with - will that work or not. I figured out the other way first, and was willing to keep it if I absolutely had to, but I don't know for sure if I prefer it or not.

So, then, I would also amend my original question, to also ask which is a better design, if they both prove to work? Keeping the multiple subnets, or pushing them all to one again. For that, I'd provided some knowledge about how my network works, and how I have to change the things at times, according to the standards.

I've almost figured it out, with the minimum help, and the troubleshooting I have done. It took this thread to make that progress though. I have to wait until I get home to do more troubleshooting, since I can't get to it from here anymore with the changes I've made.

Hopefully, this helps. It's valuable data to me, this thread, but if it's not to everyone else, I'll do something about that, while saving it for me - if my permissions allow. If I'd seen this thread before I started, and especially, mine on my site: http://smileynetmain.createaforum.co...cross-subnets/, I'd have had no trouble figuring it out WITHOUT all this work. That is the purpose for me, of being able to add documentation in this way.