A good number of private replies from people and their "day job" addresses. Most have asked for prior permission before quoting them. In general, three default-free global backbone providers stated they would love to see something like this available, from IANA is the prefered answer. Some would like to see more than just IANA address information, and other contend that would be a can of worms and opens some risk issues. It seems that there is general support and that people would use such a service if available and reliable. If you have comments on this, and can post publicly, please do. Thank you john brown speaking for himself
Cool, maybe we're making progress. The last N times this has come up, the biggest X the big IP backbones showed a distinct lack of interest or in one case extreme hostility to the idea. I've suggested an AS-NULL(AS0) or AS-RESERVED machine parsable macros for unassigned prefixes which should have no routes (including more specific routes) which could be automatically included in router configurations. Or at least queried when debuging stuff. Every network block should be assigned an "responsible party." I'm avoiding using the word "owner". By default IANA would be the responsible party for all "RESERVED" address space, and listed as such in IANA, RIR, or where ever we decied to keep the information. As address space is assigned, allocated, delegated, etc, the reserved space would be split so you can tell the difference between address squaters and valid assignments. RESERVED (Not released by IANA for use) ALLOCATED (Available for network allocations, but not in use) ASSIGNED (Assigned for use by an entity, may be routed now or soon) CONNECTED (Connected to the global Internet) MULTICAST (Not a valid source address) SPECIAL (Matians, we don't know where they come from, drop on sight) EXPERIMENTAL (Consenting parties only) PRIVATE (Local use only) On Wed, 4 Sep 2002, John M. Brown wrote:
A good number of private replies from people and their "day job" addresses. Most have asked for prior permission before quoting them.
In general, three default-free global backbone providers stated they would love to see something like this available, from IANA is the prefered answer.
Some would like to see more than just IANA address information, and other contend that would be a can of worms and opens some risk issues.
It seems that there is general support and that people would use such a service if available and reliable.
If you have comments on this, and can post publicly, please do.
Thank you
john brown speaking for himself
I'm concerned with having "to much data" in the system. This invites mistakes, potential abuse and other problems. By having only: RESERVED or ALLOCATED and having that publishd by IANA, we reduce the potential of mistakes affecting "real users". If the RIR's are going to provide more data, then they need to upgrade their business and expense models to support live people 7x24x365 so that mistakes are fixed QUICKLY. Just my own personal $.02 on the topic. I would suggest, crawl, walk, run with this idea. Lets first get IANA up and going, then see how well that works and move forward if it makes sense and the appropriate protections can be in place. john brown speaking for himself only On Wed, Sep 04, 2002 at 02:34:27PM -0400, Sean Donelan wrote:
Cool, maybe we're making progress. The last N times this has come up, the biggest X the big IP backbones showed a distinct lack of interest or in one case extreme hostility to the idea.
I've suggested an AS-NULL(AS0) or AS-RESERVED machine parsable macros for unassigned prefixes which should have no routes (including more specific routes) which could be automatically included in router configurations. Or at least queried when debuging stuff.
Every network block should be assigned an "responsible party." I'm avoiding using the word "owner". By default IANA would be the responsible party for all "RESERVED" address space, and listed as such in IANA, RIR, or where ever we decied to keep the information. As address space is assigned, allocated, delegated, etc, the reserved space would be split so you can tell the difference between address squaters and valid assignments.
RESERVED (Not released by IANA for use)
ALLOCATED (Available for network allocations, but not in use) ASSIGNED (Assigned for use by an entity, may be routed now or soon) CONNECTED (Connected to the global Internet)
MULTICAST (Not a valid source address)
SPECIAL (Matians, we don't know where they come from, drop on sight) EXPERIMENTAL (Consenting parties only) PRIVATE (Local use only)
On Wed, 4 Sep 2002, John M. Brown wrote:
A good number of private replies from people and their "day job" addresses. Most have asked for prior permission before quoting them.
In general, three default-free global backbone providers stated they would love to see something like this available, from IANA is the prefered answer.
Some would like to see more than just IANA address information, and other contend that would be a can of worms and opens some risk issues.
It seems that there is general support and that people would use such a service if available and reliable.
If you have comments on this, and can post publicly, please do.
Thank you
john brown speaking for himself
RESERVED or ALLOCATED
and having that publishd by IANA, we reduce the potential of mistakes affecting "real users".
Actually, this was part of the original RAdb. All the RESERVED space was mapped to AS-0. This was not considered useful and it was dropped. Cisco (and perhaps others), co-opted AS-0 for their own nefarious purposes.
On Wed, 4 Sep 2002, John M. Brown wrote:
I'm concerned with having "to much data" in the system. This invites mistakes, potential abuse and other problems.
By having only:
RESERVED or ALLOCATED
I'm ok with anything, as long as we try to move in the forward direction. BTW, IANA needs to remember to ALLOCATE addresses used by themselves. One problem with the current system is its difficult to tell when you have a squatter announcing a more specific block, or if it has really been allocated to them. Sean Doran demonstrated this many years ago.
and having that publishd by IANA, we reduce the potential of mistakes affecting "real users".
Actually we don't reduce the potential for mistakes. It just makes it easier to track down the culprits.
If the RIR's are going to provide more data, then they need to upgrade their business and expense models to support live people 7x24x365 so that mistakes are fixed QUICKLY.
Just my own personal $.02 on the topic.
I would suggest, crawl, walk, run with this idea.
Lets first get IANA up and going, then see how well that works and move forward if it makes sense and the appropriate protections can be in place.
Go for it. I've already submitted my recommendations on the new US national cyberprotection plan to the US Government. I don't know if they'll choose any of my ideas. I would much prefer to see a group of Internet engineers solve the problem. We've been talking about it since 1995. Instead the proposed technical solutions keep getting more and more complex to avoid dealing with the real problem. I think the actual solution is much simplier, but requires cooperation from at least the largest ISPs, RIRs and IANA. Yes, it requires more work, but its a lot less complex than some of the other ideas I've seen recently.
participants (3)
-
bmanning＠karoshi.com -
John M. Brown -
Sean Donelan