Running out of IP ranges?

If you look at IP Range Wiki Page you can see that we’ve only got 56 /24 ip ranges left (164-191 and 212-239, excluding the top range 241-247). On the [URL=“http://wiki.ctwug.za.net/w/index.php?title=CTWUG_IP_Ranges&oldid=6172”]15th April we had 113 (122-195, 208-239 and 241-247), so if we carry on at the existing rate (8.3/month), we’ll have exhausted the remaining ranges in about 6 months!

Is it time to get another /16 ?

or ipv6? :stuck_out_tongue:

1 Like

on a more serious note I have also noticed it. a Good cleanup should also help.

ive done a ping scan the other day… only 2.5K active ip’s… so theres quite a big waste!! a good cleanup will be more than enough fixing the problem

Windows 7 and Vista do not respond to pings from devices not on the same subnet unless you change the firewall rules.

I know, so double it, triple it :slight_smile: still doesnt get close to the amount ip’s issued :smiley:

I’m running into exactly the same problem at work.

How many guys have been allocated a /29 and only have one computer connected it?

Maybe we should seriously think about allocating /30’s to end users instead of /29’s.

When I joined the wug many moons ago we only got a /32 and had to NAT. Cant remember that it bothered us very much :wink:

Hehehe gooood ol’days!!

I have already reserved 172.26.0.0/16 and 172.27.0.0/16. We have the option of doing a full renumber to these two /16’s (172.27.0.0/16 is only reserved until the end of 2013 pending the return of 172.18.0.0/16).

So our options are:

  1. Renumber the entire WUG to 172.26.0.0/15 and hand back 172.18.0.0/16 (we would have a year and 3 months.

  2. Add 172.26.0.0/16 range to our IP pools and start handing out addresses from this new range. We would need to make sure routes on RBs are 172.16.0.0/12 to ensure this works.

  3. Do a cleanup and group areas into better ranges and consolidate some fragmented subsets etc.

Who wants to renumber CTWUG? :smiley:

Errrr you serious??!! :smiley:

Naturally my vote is with option two :stuck_out_tongue:

what about the good old 10 range?

10.0.0.0 / 8

If each major wug in SA takes a /14 we are set for a ver very log time.

what about the good old 10 range?

10.0.0.0 / 8

If each major wug in SA takes a /14 we are set for a ver very log time.

@Beetle001, where does one reserve private IP address ranges?

How many nodes are active? 900? That’s about 4 x Class C. Add another 4 x Class C for management and overhead. That gives us about 2000 IP addresses. Add another 50% just because we can and we’re up to 3000 IP addresses. If memory services me, a /16 is around 65,534 IP addresses. I vote for a cleanup.

So do I!! IP cleanup ftw!! Its gonna be soooo much fun:D

voting for change is pointless…instead of voting for a cleanup someone will have to volunteer to take ownership and run with it.

After taking abuse while trying to clean up a highsight, it won’t be me.

So instead of voting, rather volunteer to take the project and run with it.

We try and keep all the WUGs on seperate IP ranges, so if they ever join we don’t have to worry about conficts. This is controlled on http://www.wug.za.net/

+1

Although I’d rather see multiple high site admins taking this on. It’s a bit much for one person.

How do admins handle inactive links at the moment? Over here we monitor our links, and push a use it or loose it policy. If there are wuggers who switch their links off for long enough and ignore our notices, we revoke their IP addresses and return them to our pools. When you’re not charging money for a service you have to implement a use it or loose it strategy, it’s that simple. All admins should be doing this in their districts/areas.

Ping scans aren’t very useful for the reason pmurgs pointed out, because of the inefficient nature of subnetting, and because of the way IP addresses are assigned and used. There’s no way anyone can realistically expect most or even half of all usable address space to be actively in use - networks simply don’t work like that. For example, I have a /28, but right now as I type, less than half of it is live. At the same time there’s no way my network would be usable with a smaller allocation, because the next size down is less than half the size, and because the number of devices on my network is dynamic, not static. If you think in terms of individual IP addresses you have the wrong perspective of IP networks. Think in terms of IP subnets, and make sure there are no prefixes routed to networks that are inactive.

As Beetle pointed out, we have no shortage of IP addresses. Reducing the size of allocations to wuggers would be a painful waste of time. If anything allocations should be larger. Who has less than 5 devices on their networks these days? What are we, Amish? Personally I have 5 devices all to myself on my network, that excludes the shared devices like switches, routers, servers, TVs, etc. I rarely give less than a /28 to wuggers, and I always reserve the adjacent /28 if it’s available in the pool. High sites get no less than a /24, and also get reserved the adjacent /24 if it’s available in our pool.

So no shortage of IP addresses, but what feels like a mess of random, unplanned allocations. Why are there prefixes smaller than /25 on the IP Address wiki page?? Even some /29s there, geez! Time to start putting the district model to better use. Search the wiki for “deepsouth” and see how we’re doing things… (we’re still busy pulling up our broeks)

So don’t worry about running out of addresses. Do worry about how you plan and make your allocations, and do worry about monitoring your networks for dead links. Once we can get that right we’ll hopefully be ready for the next big step - IPv6. :slight_smile: