Mine is 1 for router/gateway, 2-199 is static with 2-9 (technically 1-9) being network gear, 10-19 being servers/related, 20-29 computers, 30-39 phones… all the way up to 100-119 for cameras and access, currently. And then 200+ for dhcp.
Everything on the network runs dhcp but gets a static ip from the router. If I ever forget to assign a new device, it’s obvious since it won’t be near the related equipment, and it’ll be like 213 or something, sticking out like a sore thumb.
I set the gateway to .1 and everything up to 100 I move out of DHCP into static pool for some nodes (speakers, servers, gateways, SIP phones, access points, switches, printers, vacuum cleaners, etc.). Simply because in this case, you have to type fewer digits to log into these boxes.
Edit: Yes, 255 is technically not ‘reserved’ by itself. I was only thinking about the addresses scoped to /24 for a simple answer. A broadcast address is reserved, though.
Well technically a 10. network could be a /8 if you really wanted to. But in this case the /23 was specified above. I don’t think it’s necessary but that’s what op was talking about and the 212.255 is a valid address, as is 213.0
Yeah that’s fair. Maybe OP is planning a nonsense amount of IoT devices and cameras? or they just want to use some unorthodox IPs like .255 or .0 for fun 🤷
Thanks for the correction on the the broadcast address, btw. The details are important.
At home, I have 172.16/12 scoped out for my cluster subnet and just use the second and third octet to denote service function. It’s excessive scoping, but it’s an internal scope for a private network and broadcasts are generally contained to the cluster network. It would be trivial to split that into a 172.16/16 if I ever had the desire. (Speaking bluntly, it really doesn’t matter much in this case. I’ll need to split it out to keep Longhorn traffic segregated later, but it’s not an issue now.)
The rest of the network is reasonably scoped out across 192.168 ranges and chopped up across a few VLANs.
My addressing scheme is split into “public” and “stay the fuck out”, with most 172 addresses being the latter. At home this is great. In a large network, scoping out such a large 172 subnet could be detrimental. (“It depends” on the detrimental bit and requires proper planning and understanding about why subnets were even built into IPv4 from the get-go.)
Edit: Most addressing is static on my cluster network, with DHCP only handing out addresses based on MAC address. The 172 subnet has very restricted dynamic IP ranges.
Regardless of what is what, it’s a scheme I recognize and that is the more important bit.
It might also make you really grumpy if you decide to use a VPN or containers - those often use 10/24 subnets and having an overlap with your LAN can make things spicy.
I could see something like /20 being useful not so much for the number of addresses but to group things by use case, then /8 because you don’t care and it doesn’t matter. Also 10.0 is quicker to type and cleaner to look at than 192.168
It makes sense if you think about it, but seeing IPv6 addresses in the IPv4 routing table still makes professional network engineers stop and boggle for a moment
.254
Anything else is mental illness. .1 is sus but acceptable in a pinch. But why cut into the DHCP pool like that. 1-100 is DHCP, obviously.
My DHCP is 100-something I dont remember because I don’t care about or respect the devices using DHCP.
As God intended. If you’re important in any context, you have a static IP. Otherwise you’re a second or third class device.
Gonna use 10.233.212.254 on my 10.233.212.0/23 👍
There are no windows in your house.
https://vlipsy.com/clips/dr-tran-there-are-no-windows-in-your-house-aoFRHS2T
I never run into Dr.Tran in the wild and it’s a shame. More people need to support red blooded American secret agents that also have a doctorate
Buttery what!? Smoked what!? So crazy!
My soho setup is:
1-100 = DHCP
101 = Default gateway
102 - 199 = Unused
201-254 = Static, manually assigned IPs (if any)
I came up with it before I got my Net+. I told myself I’d change it as soon as it caused any problems. That was 12 years ago.
I use the low numbers for statics, less to remember/communicate if its a 1/2 digit number instead of 3.
I’m stealing this config, I love it.
Mine is 1 for router/gateway, 2-199 is static with 2-9 (technically 1-9) being network gear, 10-19 being servers/related, 20-29 computers, 30-39 phones… all the way up to 100-119 for cameras and access, currently. And then 200+ for dhcp.
Everything on the network runs dhcp but gets a static ip from the router. If I ever forget to assign a new device, it’s obvious since it won’t be near the related equipment, and it’ll be like 213 or something, sticking out like a sore thumb.
.1 gateway .2-.99 static range .100-.254 dynamic range (dhcp)
btw. /23 networks are cool, too.
yes to /23. Lower 24 is all static. Upper 24 is all DHCP. And the GW lives right in the middle at .1
DHCP first is unhinged bruh
1-127 is static, DHCP starts at 128.
Libvirt/QEMU agrees with me so I must be right
I set the gateway to .1 and everything up to 100 I move out of DHCP into static pool for some nodes (speakers, servers, gateways, SIP phones, access points, switches, printers, vacuum cleaners, etc.). Simply because in this case, you have to type fewer digits to log into these boxes.
Why not 255?
255 is reserved for broadcast.
Edit: Yes, 255 is technically not ‘reserved’ by itself. I was only thinking about the addresses scoped to /24 for a simple answer. A broadcast address is reserved, though.
When in doubt: https://www.calculator.net/ip-subnet-calculator.html
10.233.213.255 is broadcast, 10.233.212.255 is a usable address …
How big of an IP scope do you use at home?
Do you really need something larger than a /24?
I’m quite fond of 10.0.0.0/8. That gives you a lot of wiggle room, just in case.
Just a little bit of wiggle room…
It’s useful to keep the 16 million IoT devices you have running on WiFi segregated a bit.
Well technically a 10. network could be a /8 if you really wanted to. But in this case the /23 was specified above. I don’t think it’s necessary but that’s what op was talking about and the 212.255 is a valid address, as is 213.0
Sure, you could. But you’d be going out of your way to setup an internal IP scope you’ll never fully utilize.
I suppose it doesn’t really matter, but it just feels wrong.
Yeah that’s fair. Maybe OP is planning a nonsense amount of IoT devices and cameras? or they just want to use some unorthodox IPs like .255 or .0 for fun 🤷
Thanks for the correction on the the broadcast address, btw. The details are important.
At home, I have 172.16/12 scoped out for my cluster subnet and just use the second and third octet to denote service function. It’s excessive scoping, but it’s an internal scope for a private network and broadcasts are generally contained to the cluster network. It would be trivial to split that into a 172.16/16 if I ever had the desire. (Speaking bluntly, it really doesn’t matter much in this case. I’ll need to split it out to keep Longhorn traffic segregated later, but it’s not an issue now.)
The rest of the network is reasonably scoped out across 192.168 ranges and chopped up across a few VLANs.
My addressing scheme is split into “public” and “stay the fuck out”, with most 172 addresses being the latter. At home this is great. In a large network, scoping out such a large 172 subnet could be detrimental. (“It depends” on the detrimental bit and requires proper planning and understanding about why subnets were even built into IPv4 from the get-go.)
Edit: Most addressing is static on my cluster network, with DHCP only handing out addresses based on MAC address. The 172 subnet has very restricted dynamic IP ranges.
Regardless of what is what, it’s a scheme I recognize and that is the more important bit.
It might also make you really grumpy if you decide to use a VPN or containers - those often use 10/24 subnets and having an overlap with your LAN can make things spicy.
I could see something like /20 being useful not so much for the number of addresses but to group things by use case, then /8 because you don’t care and it doesn’t matter. Also 10.0 is quicker to type and cleaner to look at than 192.168
The highest address of a subnet is the broadcast address. You can’t assign it to a host.
But 10.233.212.225 is middle of the range for the /23.
I had students like you when teaching Cisco classes…
Their peers heads would always explode, lol.
If you use a /23 you can use the middle .0 too.
My favourite dumb networking trick though is that Linux lets you use an IPv6 address as an IPv4 gateway.
That works so long as the gateway is running dual stack.
But why can’t I use IPX in my routing table for IPv4???
Thanks I hate it.
Makes sense, the gateway is never a target of the ip packet, so it’s really only used to find the MAC address of the gateway.
It makes sense if you think about it, but seeing IPv6 addresses in the IPv4 routing table still makes professional network engineers stop and boggle for a moment