[ietf78-tech] Impact of long leases on IPv4 DHCP address ranges for wireless

[Rob Nagy]

A side note. I had an email exchange with Alan DeKok today(missed his call). He has loaded far more users then we are planning with no ill effects. The issue will be RAM if there is any issue. I generated 10,000 random user/pass combos. I will test it out in my lab before touching the VMs, just to see the effects. And I will ask before  touching VMs just in case it doesn't go well. 

Alice

Sent from my iPhone

On Jul 8, 2010, at 10:35 PM, John Kemp <kemp at network-services.uoregon.edu> wrote:

> I'll add a supporting point here.  Since we have ip+Mac already for v4, switching the insert rule to ipsets ( hashing) becomes easy.  The long v6-only-has-Mac chain is still there, but with connection tracking, a lot of that will be hashed as well.
> 
> so I'm mildly optimistic as is.
> /jgk
> 
> Sent from my iPad
> 
> On Jul 8, 2010, at 7:20 PM, Randy Bush <randy at psg.com> wrote:
> 
>>> Uh... I thought the design goal was: login either 1 time , or at worst
>>> 1 time a day.
>> 
>> dhcp leases renew automagically, yes?  lease renewal does not require
>> re-auth.  well, should not.
>> 
>> but i do not think it wise to make folk auth more than once a day as
>> long as they stay in range of one of our APs
>> 
>> randy
> _______________________________________________
> ietf78-tech mailing list
> ietf78-tech at daedelus.com
> http://www.daedelus.com/mailman/listinfo/ietf78-tech