So the company I work for does a lot of property-wide wifi installs. Been doin it for a few years. We just finished two buildings and are working on a third for this particular network that will total somewhere around 900 units. The guy "In charge" of this project is an IT guy and he got pulled away today. I overheard him say that they were running out of IP addresses next door.

I didn't go to school for networking/IT so I have to pick up all of that stuff on my own. So, can you guess what my question is?

What exactly will they be doing to increase the number of IP addresses they can dish out tonight? (They're doing it around 2 a.m. so as not to disrupt people's service.) Is it something as simple as changing the DHCP range in the router? This entire network is guest based since these are at least 90% rental units (tourist town, vacation haven) so they're all using the same SSID and password. So, is it that simple?

It may be pertinent to know, so, as part of our rack we're using a Ruckus Zone Director (and Ruckus waps of course), something called a "slice" that I have no idea what it does really (If you know what it does, feel free to explain!) and....aside from a router, that's all I can think of that might affect it? If there's something else, please let me know.

So yeah, what ya got?

Thanks!

You can increase size of subnet to /23 or /22 and increase the available address space in a single domain (riskier), or you can create additional subnets (VLANs) with their own individual address space. The first option is risky because as the broadcast domain size increases so does the aggregate broadcast and multicast traffic--especially risky on a BYOD network like you've described. VLANs are more complicated to configure and manage, especially if the use case requires significant interVLAN routing. That puts additional processing burden on whichever device(s) are responsible to handling that traffic (router, L3 switches, etc.)

Whenever I hear slicing mentioned it's generally in reference to cellular network virtualization. I'm not sure what they're referring to in your case. WiFi airtime fairness maybe?

All 900 units are using one router ?

I would think there multiple routers and multiple v LANs to achieve enough ip addresses for 900 units.

Assuming 900 units could need over 3000 ip addresses

He meant eunuchs not units. They don't have the balls to put one router in 900 units :-D

How does that even work? 900 units share a single internet access account?

It's not exactly a rare situation guys. There are plenty of corporate offices, hotels, institutional campuses, government facilities, data centers, etc. that deal with this situation every day. I'm pretty confident the backbone of this particular network isn't a Asus all-in-one from Best Buy.

Sean,
but do you have an answer or any advice for Grasshopper?

Sean explained it well -- very well in fact.

Grasshopper, with all due respect, this is not something you'll pick up from watching a YouTube video or by getting an answer on RemoteCentral forum.

I've been messing enough with networks to know what my limits are.
While I use few subnets in my own house because of my 300+ IP devices, several DHCP pools, etc., there is no way that I would just 'take over' what you're describing. One wrong command or inter VLAN route and you could crash the network or worst, open it up to a hacking attack.
900 units = 900+ pissed off people when they can't get on the net.

See the the very first reply in this thread.

What i've learned from this thread:

Mario lives in a palace.

Can jump to a class B network and have 65,536 available ip's.

I recommend a linksys wrt54g to handle all those.

Jk. Lol.


I guess a tad off topic but in reference to my original reply Classful networks are an interesting story though... The huge gaps between the available iPs in class c and b networks actually led to cidr or classless networks. Basically they were running out of ips. 254 wasn't enough in class c for many companies. Class B wasted too many ip's. So the current classless method was created. Here is more info:

Full article: [Link: tools.ietf.org]

1. Exhaustion of the Class B network address space. One fundamental
cause of this problem is the lack of a network class of a size
that is appropriate for mid-sized organization. Class C, with a
maximum of 254 host addresses, is too small, whereas Class B,
which allows up to 65534 host addresses, is too large for most
organizations but was the best fit available for use with
subnetting.


You can learn tons by just googling classful classless networks and googling subnetting but I'm sure it can be overwhelming for many.


[Link: networkingconcepts4u.blogspot.in]

+1! lol

gerard - I did google some stuff after reading this thread and even the 'basic' stuff got a little weird for me. Subnets and CIDR notation...yeeaah...I'll be reviewing those.

If all of them are behind some sort of super-NAT type router, couldn't you just use IPV6 addresses internally? At least with Windows all machines by default (unless you disable it) automatically get an internal, unique IPV6 address.

As long as the router in that situation allowed for a NAT'ing of internal IPV6 to external IPV4 it wouldn't matter that you were using IPV6 internally.

Or are all of these IPs you are passing out actually public, routable IPs?If that's the case then above would be useless.

But, within a LAN, IPV6 makes perfect sense and is stupidly easy to set up since a default address can be created from the machine's MAC address, no need for any DHCP to hand out unique addresses. It's kind of crazy that so much work as been put into IPV6 and it still gets used so little.

All I know is that I use the forklift for a lot more than I thought I would.