Advertisement
Not a member of Pastebin yet?
Sign Up,
it unlocks many cool features!
- Van: Bram Peeters <Bram.Peeters@surfnet.nl>
- Datum: donderdag 26 september 2013 13:17
- Aan: "org-l@ams-ix.net" <org-l@ams-ix.net>
- CC: "christian.kaufmann@ams-ix.net" <christian.kaufmann@ams-ix.net>
- Onderwerp: AMS-IX plans, concerns, and why we will vote "no"
- Dear members, board -
- We'd like to explain our view on the discussion, and why we will vote "No". Three main reasons:
- - What is the plan, why the "need"
- - Legal uncertainties
- - Working with members
- Our (admittedly late, but we were reviewing the - legal - documentation) suggestion would be to not have this extraordinary meeting at all, and have a proper, well-prepared and well-informed discussion + vote in a standard GM.
- In future discussions we would appreciate interactions with the members that show a true appreciation of their contributions, engagement and concerns; for example being transparent, timely, clear and open about plans with strategic scope such as the current item under discussion.
- What is the plan, why is there a "need"?
- ====================================
- - Let's remember that the value of an exchange resides in the quality and number of connectors that are present. The AMS-IX in Amsterdam is a brilliant exchange. How much value (and risk) is effectively added by growth in other locations is debatable, should perhaps indeed be debated. How that would work for the Open-IX plans is not clear at all.
- - The current expansion plans at the AMS-IX seem to be quite strategic (considerable commitment/investment - including management attention - and hard to turn back). With the lack of proper argumentation we fail to understand the value for the member base, although there are obvious risks involved.
- - The operational model is unclear and seems to be different from the models discussed in the LTCS (collaborative partnerships, where partners assumedly pick up the main risk - and with the board to judge that on behalf of the GM). Whether the current plan is indeed "start small, make things easy, scale fast (or fail fast with limited risk)" is unclear.
- - The LTCS therefore does not necessarily support these plans. The argument that has repeatedly been used that (paraphrased) "this is all part of the LTCS, just help us to clear this practical issue" is not as valid as it may look.
- - Where previously the argument has been that the AMS-IX can/should play a role in an upcoming Internet market, this can hardly be said about the US. For the Open-IX initiative many other ways to 'develop our successful model in other places in the world' could be chosen, as has been pointed out in a few mails. For example teaming up with DE-CIX and LINX to avoid unnecessary competition between European exchanges on US soil. The model does not necessarily need to be executed by the AMS-IX.
- - Does it work? In HK - some two years after starting operations - there are now 10 connectors with approximately 3Gbps traffic, perhaps too early to really evaluate but at the moment this is obviously still a long way from the success that clearly justifies AMS-IX's involvement. The CAR-IX (with 11 connectors and some 4 Gbps traffic) could indeed be qualified as a "for the good of the internet" project. Obviously both are in line with "start small, make things easy, scale fast (or fail fast with limited risk)".
- Legal uncertainty
- ===============
- Like many other participants SURFnet has considerable concerns regarding the US regulations on the interception of communication, commonly called “the Patriot Act”. We have read the memorandum on this subject of 24 September 2013.
- We feel this memorandum is too little, too late. Too little, because the memorandum itself is not necessarily convincing; there is considerable room for uncertainty and quite a few conclusions it draws are not sufficiently supported with arguments to assess their validity.
- The interpretation of data being under the custody or control depends heavily on the legal structure chosen between the BV and the US entity, which is not discussed in the memorandum. Furthermore we think the matter of jurisdiction is of great importance, the memorandum does not give us the arguments to feel assured that any structure exists where the US government will not claim jurisdiction.
- The memorandum notes that members are already subject to information requests in Europe. This is obviously true, but if data is collected under US law there is only a very limited protection for foreign users as constitutional safeguards are not applicable (this is also discussed in a study we commissioned in 2012: http://www.ivir.nl/publications/vanhoboken/Cloud_Computing_Patriot_Act_2012.pdf, also picked up by CBS and other media channels: http://www.cbsnews.com/8301-205_162-57556674/patriot-act-can-obtain-data-in-europe-researchers-say/)
- The memorandum explicitly leaves room for debate, was simply published too late to properly think through and discuss its argumentation.
- Working with members
- ====================
- - Communication: we are quite disappointed with the way this GM and the current discussion has been handled by the board. The initial invitation to the GM was - least to say - not very informational, given the potential impact. An initial mail by Erik Bais - with quite a few valid and important points - remained unanswered for close to two weeks, leaving the discussion to one week before the GM. After his re-send got a considerable number of +1's, some comments from the board and management started to trickle through and some additional information was presented. We do support the view of others that if this subject is sufficiently important to require a GM to vote on it that the GM should be provided _on time_ with sufficient qualitative and quantitative information.
- - Quality of documentation: the document 'US resolution reasoning' provides a sketchy interpretation of _a_ global strategy that can not be found in the LTCS (although it is presented as such). The document then quotes a few lines of the LTCS and the Articles that in our opinion should not be interpreted as 'very strong support for AMS‐IX to develop services in a number of the locations Open‐IX has defined'. Next, the reasoning proceeds to pointing at the Jones-Day memorandum that – as said - does not necessarily support the conclusions the board seems to draw. Last night a Q&A was published that, next to addressing some other issues, uses the same approach in presenting a strategy that seems to be based on the LTCS but that is in fact not necessarily supported by that LTCS.
- - Understanding the concerns: even though quite a few members asked for more information and supporting evidence/argumentation the contributions from the board have been limited. The discussion now seems to focus on the privacy issues, probably because it is the only subject where the board and management have added some actual input whereas other questions have remain unanswered or are removed by the Authority of the LTCS . The Board also seems to have missed the point that quite a few members really want to have this discussion; the changes to the proposed resolutions do not accommodate this - the Chair of the Board explicitly stated that the discussion is only centered on the legal aspects. It should not only have been.
- - the LTCS seems to be used as the authority to do away with all "why" questions ("members voted on the LTCS (unclear how many), the LTCS covers these plans, the board therefore has the remit to decide, however we just need you to approve some technicalities"). Admittedly, the LTCS has over the past few years been incorporating more "expansion" plans, and in communications the AMS-IX management has been putting more emphasis on this - however, the subject of expansion has never been properly presented, the discussions that were related have been limited and were at the same time not trivial (the discussion on the ill-fated services BV, the discussion in GM32 on Reseller-plus less than two years ago etc). This is a discussion that should be brought to the GM.
- Kindest regards -
- Bram
- --
- Bram Peeters
- Head of Network Services
- SURFnet bv - http://www.surfnet.nl/bram.peeters@surfnet.nl
- tel: +31 302 305 305
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement