Re: ARIN, was Re: 72/8 friendly reminder
From: Michael.Dillon Date: Thu Mar 24 11:34:52 2005
--------------------------------------------------------------------------------
The other consequence is that the membership takes on the responsibility for ARIN's actions. Not the staff's actions, but ARIN's actions. If there is any dysfunction in ARIN, I suspect that it lay here.
Yes, this is what I believe. The ARIN membership is more passive than I think is healthy for the organization. Thus, the organization is dysfunctional.
I agree, I'd certainly like to see more people actively participate in the process. If nanog folks believe that the ARIN membership is not getting the right stuff done... How do we fix this problem? How do we get more operators involved and active in the RIRs? I think colocating 1 ARIN meeting/per year with Nanog in the fall has been a help. ARIN isn't perfect but it could be a lot worse. In some ways I think the issue you describe is an industry wide problem. There are many different groups (RIRs, ICANN, IETF, Nanogs, etc...) and participating in all of them is a lot of effort, especially when most of us already have full-time jobs. We could of course create a huge beuarcratcy with lots of people to study the issues and make policy, but that hasn't been the way the Internet has developed and is counter to what many operators think is best for the Internet. That also requires money. Is that what people want? I don't think so, but I could be wrong. Andrew (also a member of the ARIN Advisory Council)
At 17:01 +0000 3/24/05, Andrew Dul wrote:
I agree, I'd certainly like to see more people actively participate in the process. If nanog folks believe that the ARIN membership is not getting the right stuff done... How do we fix this problem? How do we get more operators involved and active in the RIRs?
In the spirit of cart and horse, it's not about getting more operators involved in ARIN. It's about getting operators to use ARIN as a resource in the proper way. (I'm addressing operators here as this is NANOG.) What do operators expect from ARIN? Most ARIN policies are centered on the administrative function of allocation of address space and AS numbers. Is that all there is? Are the existing policies all that are needed? Are there concerns about the live-in-the-network registry services like WhoIs, DNS, IRIS, routing registry? There are not many policy proposals (lame delegations, privacy concerns with WhoIs) in play covering operational considerations. ARIN staff has begun work on documenting the registry service level agreements, there was a presentation on this in October. There has been little discussion on this by anyone since the presentation. If WhoIs is out, reports fly on NANOG. But has anyone ever tried to quantify what level of service is expected of ARIN's computing facilities? If the staff is doing a good thing by documenting SLA's, then they should be encouraged to continue. There is routing security research work that would require the RIR's to issue certificates for use in route update validation. I would hope that someday, before anything goes live, there are operator-led tests involving support from ARIN.
I think colocating 1 ARIN meeting/per year with Nanog in the fall has been a help.
I would caution that "attending meetings" is neither a sign of contribution nor a sign of progress. Don't get me wrong, making meetings easier to attend is good, but we shouldn't attend meetings because it is easy, fun or entertaining. I prefer to have fun at home.
ARIN isn't perfect but it could be a lot worse. In some ways I think the issue you describe is an industry wide problem. There are many different groups (RIRs, ICANN, IETF, Nanogs, etc...) and participating in all of them is a lot of effort, especially when most of us already have full-time jobs.
Participating in all of them *is* a full-time job. ;)
We could of course create a huge beuarcratcy with lots of people to study the issues and make policy, but that hasn't been the way the Internet has developed and is counter to what many operators think is best for the Internet. That also requires money. Is that what people want? I don't think so, but I could be wrong.
One the one hand, what built the Internet isn't what will maintain it. A bureaucracy will be needed, the challenge isn't to prevent it but to build the best one possible. If ARIN goes unchecked it'll either be a weakened organization unable to serve the community (chaos ensues) or it will become an ogre, burdening the community (suffocation). It benefits operators to be involved, but the real trick is to realize what kind of involvement is needed. -- -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Edward Lewis +1-571-434-5468 NeuStar Achieving total enlightenment has taught me that ignorance is bliss.
--On Thursday, March 24, 2005 3:20 PM -0500 Edward Lewis <Ed.Lewis@neustar.biz> wrote:
At 17:01 +0000 3/24/05, Andrew Dul wrote:
I agree, I'd certainly like to see more people actively participate in the process. If nanog folks believe that the ARIN membership is not getting the right stuff done... How do we fix this problem? How do we get more operators involved and active in the RIRs?
In the spirit of cart and horse, it's not about getting more operators involved in ARIN. It's about getting operators to use ARIN as a resource in the proper way. (I'm addressing operators here as this is NANOG.)
Fair enough...
What do operators expect from ARIN? Most ARIN policies are centered on the administrative function of allocation of address space and AS numbers. Is that all there is? Are the existing policies all that are needed?
Other than a community service/educational role on issues related to the above policies, yes, that is the limits of ARIN's charter. Other issues are the purview of ICANN, IETF, IESG, and ISOC.
Are there concerns about the live-in-the-network registry services like WhoIs, DNS, IRIS, routing registry? There are not many policy proposals (lame delegations, privacy concerns with WhoIs) in play covering operational considerations.
There are not many such proposals in play at the moment because the ARIN community reached consensus around most of these issues over the last two years. There seems to be general agreement that the current state of things is acceptable in terms of Whois and DNS. While ARIN runs a Routing Registry as part of it's public service focus, I do not believe that ARIN should have a defining role in the IRR process. In general, that also is the purview of the IETF.
ARIN isn't perfect but it could be a lot worse. In some ways I think the issue you describe is an industry wide problem. There are many different groups (RIRs, ICANN, IETF, Nanogs, etc...) and participating in all of them is a lot of effort, especially when most of us already have full-time jobs.
Participating in all of them *is* a full-time job. ;)
Right, but, the portion of the internet community which consists of organizations willing to pay an FTE to do that job is very small. Owen -- If it wasn't crypto-signed, it probably didn't come from me.
At 13:01 -0800 3/24/05, Owen DeLong wrote:
There are not many such proposals in play at the moment because the ARIN community reached consensus around most of these issues over the last two years. There seems to be general agreement that the current state of things is acceptable in terms of Whois and DNS. While ARIN runs a Routing Registry as part of it's public service focus, I do not believe that ARIN should have a defining role in the IRR process. In general, that also is the purview of the IETF.
Here's my dilemma. On the one hand I hear calls for greater operational input to ARIN. On the other hand is empirical evidence that there isn't much input being given. What I have been trying to do extract what latent operational input might be fed to ARIN, judging from discussions I have seen at other RIRs, the IETF, etc. If there aren't follow ups to these ideas, then I would conclude that ARIN isn't dysfunctional and is operating as it should be, an idea supported by what is above. If there are ideas forthcoming, then maybe there is a need to encourage participation. This thread was ignited by the desire to have a pingable address in newly allocated blocks (from IANA to ARIN), and maybe Randy's suggestion is all that is needed - simply asking ARIN to do this. Maybe policies aren't the only way to influence ARIN's operation. -- -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Edward Lewis +1-571-434-5468 NeuStar Achieving total enlightenment has taught me that ignorance is bliss.
Here's my dilemma. On the one hand I hear calls for greater operational input to ARIN. On the other hand is empirical evidence that there isn't much input being given.
Correct... Generally, you hear those calls coming from ARIN because ARIN is trying to maximize the involvement of its constituency. This is a good thing, but, should not create the illusion that there is not already significant involvement. This is sort of one of those "We can always do better" kind of issues, and, I think that active solicitation is better than the alternatives.
What I have been trying to do extract what latent operational input might be fed to ARIN, judging from discussions I have seen at other RIRs, the IETF, etc. If there aren't follow ups to these ideas, then I would conclude that ARIN isn't dysfunctional and is operating as it should be, an idea supported by what is above. If there are ideas forthcoming, then maybe there is a need to encourage participation.
Got it. Yes, I think that there needs to be encouragement for ideas to be forthcoming whether such ideas exist or not. I think ARIN is doing a pretty good job of providing that encouragement.
This thread was ignited by the desire to have a pingable address in newly allocated blocks (from IANA to ARIN), and maybe Randy's suggestion is all that is needed - simply asking ARIN to do this. Maybe policies aren't the only way to influence ARIN's operation.
Right... So, things divide into two categories... Major Undertakings and changes to existing policy... Requires policy process. Easily implemented obvious wins for everyone (a pingable address within a new block would be an example here) where the first step should be a polite "Hey ARIN Staff, can this be done?" If the staff says "Sure...Easy... look for an announcement soon.", then my experience has been they tend to get implemented fairly quickly (I believe this is what I just saw from Leslie a couple of minutes ago on this very issue). If the staff says no, they generally provide reasons and suggestions. In this case, either the policy process or an alternative solution is probably in order. With a minimal reading of the policy manual and some thought, I think it's fairly easy to sort out which type of request fits in which category. If in doubt, ask the staff first, they'll be happy to tell you whether it requires policy or can be done at the staff level. Owen
I agree, I'd certainly like to see more people actively participate in the process. If nanog folks believe that the ARIN membership is not getting the right stuff done... How do we fix this problem? How do we get more operators involved and active in the RIRs?
I'd like to point out that ARIN policy is _NOT_ controlled by ARIN membership. While the ARIN BOT has final approval/disapproval authority over proposed policies, this is akin to a presidential VETO. The ARIN AC has the primary role in policy development and responsibility for judging community consensus around policies. The ARIN AC is elected by the ARIN membership, but, ARIN membership is not a requirement to run for or be elected to the AC. Further, policy proposals may be made by any member of the community, not just ARIN members. I have been an active participant in ARIN for several years now, and, only for part of that time was I affiliated with an ARIN member. In fact, I ran for AC while I was not an ARIN member. I came within a few votes of being elected. I will run again this year. It is unlikely that I will be an ARIN member when I do.
I think colocating 1 ARIN meeting/per year with Nanog in the fall has been a help.
Yes. Personally, I think ARIN is not all that dysfunctional. I think it is a lot less dysfunctional than IETF at this point. Owen -- If it wasn't crypto-signed, it probably didn't come from me.
participants (3)
-
Andrew Dul -
Edward Lewis -
Owen DeLong