[address-policy-wg] 2011-05 New Policy Proposal (Safeguarding future IXPs with IPv4 space)
- Previous message (by thread): [address-policy-wg] 2011-05 New Policy Proposal (Safeguarding future IXPs with IPv4 space)
- Next message (by thread): [address-policy-wg] 2011-05 New Policy Proposal (Safeguarding future IXPs with IPv4 space)
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Andy Davidson
andy at nosignal.org
Wed Oct 26 10:18:58 CEST 2011
On 25 Oct 2011, at 19:06, Remco Van Mook wrote: > With some first hand experience in setting up Exchanges, I would suggest > the following: > - a new IXP gets a /24 > - an existing IXP that runs out of its /24 (whether from this /16 or > another range) gets a /23 and has to return the old /24 > - an existing IXP that runs out of its /23 (whether from this /16 or > another range) gets a /22 and has to return the old /23 > (Yes, this means renumbering every single time. Yes, it's a pain. Yes, if > you run out of a /22 for your peering LAN you're fresh out of luck.) This is sane logic, but it feels very much like userland^Wimplementation documentation rathe than policy to me. Agree ? I don't mind it being in the policy though as it sounds very sane. > And while we're at it, I would suggest to add to 5.6.4: > > c. Any address space that is returned by an IXP will be added to the > reserve as outlined in 5.6.2. Agreed. Andy
- Previous message (by thread): [address-policy-wg] 2011-05 New Policy Proposal (Safeguarding future IXPs with IPv4 space)
- Next message (by thread): [address-policy-wg] 2011-05 New Policy Proposal (Safeguarding future IXPs with IPv4 space)
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]