Re: RFQ: IP service in U.K. for U.S. hosting company
>We've determined that the best option for our problem is not > peering, but the purchase of service from a provider with a good network > in the U.K. and on the continent. We'd like to receive proposals based > on this requirement. Then you should write up an RFP and send it to companies that meet your requirement. Otherwise you risk only getting quotes from people who troll the mailing lists, desperate for business. --Michael Dillon
Re: Network end users to pull down 2 gigabytes a day, continuously?
> > Then why can't they plug in Power, TV & phone line? That's > > where IPTV STBs are going... OK, I can see that you could use such a set-top box to sell broadband to households which would not otherwise buy Internet services. But that is a niche market. > Especially as more and more ISPs/telcos hand out WLAN boxen of various > kinds - after all, once you have some sort of Linux (usually) > networked appliance in the user's premises, it's quite simple to > deploy more services (hosted VoIP, IPTV, media centre, connected > storage, maybe SIP/Asterisk..) on top of that. He didn't say that his STB had an Ethernet port. And I'm not aware of any generic Linux box that can be used to deploy additional services other than do-it-yourself. And that too is a niche market. Also, note that the proliferation of boxes, each needing its own power connection and some place to sit, is causing its own problems in the household. Stacking boxes is not straightforward because some have air vents on top and others are not flat on top. The TV people have not learned the lessons of that the hi-fi component people learned back in the 1960s. --Michael Dillon
Re: Network end users to pull down 2 gigabytes a day, continuously?
> How many channels can you get on your (terrestrial) broadcast receiver? There are about 30 channels broadcast free-to-air on digital freeview in the UK. I only have so many hours in the day so I never have a problem in finding something. Some people are TV junkies or they only want some specific content so they get satellite dishes. Any Internet TV service has a limited market because it competes head-on with free-to-air and satellite services. And it is difficult to plug Internet TV into your existing TV setup. --Michael Dillon
What comes AFTER YouTube?
> Not only does this type of programming require real-time > distribution, as these shows are quite often cheaper to produce than > pre-recorded entertainment or documentaries they tend to fill a large > portion of the schedule. And since there are so many of these reality shows in existence and the existing broadcast technology seems to perfectly meet the needs of the show producers, what is the point of trying to shift these shows to the Internet? If it ain't broke, don't fix it! I do believe that the amount of video content on the Internet will increase dramatically over the next few years, just as it has in the past. But I don't believe that existing video businesses, such as TV channels, are going to shift to Internet distribution other than through specialized services. The real driver behind the future increase in video on the Internet is the falling cost of video production and the widespread knowledge of how to create watchable video. Five years ago in high school, my son was taking a video production course. Where do you think YouTube gets their content? In the past, it was broadband to the home, webcams and P2P that drove the increase in video content, but the future is not just more of the same. YouTube has leveraged the increased level of video production skills in the population but only in a crude way. Let's put it this way. How much traffic on the net was a result of dead-tree newpapers converting to Internet delivery, and how much was due to the brand-new concept of blogging? --Michael Dillon
Re: Network end users to pull down 2 gigabytes a day, continuously?
> I remember the times when I could watch mexican tv transmitted from a > studio in florida. If it comes from a studio in Florida then it is AMERICAN TV, not Mexican TV. I believe there are three national TV networks in the USA, which are headquartered in Miami and which broadcast in Spanish. --Michael Dillon
RE: Network end users to pull down 2 gigabytes a day, continuously?
> Bring that box to the living room in an attractive package and > the stats will be very different. This kind of box is very popular in England. It is called a digital TV receiver and it receives MPEG-2 streams broadcast freely over the airwaves. Some people, myself included, have a receiver that with a hard disk that allows pausing live TV and scheduling recording from the electronic program guide which is part of the broadcast stream. Given that the broadcast model for streaming content is so successful, why would you want to use the Internet for it? What is the benefit? --Michael Dillon
Re: Network end users to pull down 2 gigabytes a day, continuously?
> But what happens when 5% of the paying subscribers use 95% of the existing > capacity, and then the other 95% of the subscribers complain about poor > performance? "Capacity" is too vague of a word here. If we assume that the P2P software can be made to recognize the ISP's architecture and prefer peers that are topologically nearby, then the issue focuses on the ISP's own internal capacity. It should not have a major impact on the ISP's upstream capacity which involves stuff that is rented from others (transit, peering). Also, because P2P traffic has its sources evenly distributed, it makes a case for cheap local BGP peering connections, again, to offload traffic from more expensive upstream transit/peering. > What is the real cost to the ISP needing to upgrade the > network to handle the additional traffic being generated by 5% of the > subscribers when there isn't "spare" capacity? In the case of DSL/Cable providers, I suspect it is mostly in the Ethernet switches that tie the subscriber lines into the network. > The reason why many universities buy rate-shaping devices is dorm users > don't restrain their application usage to only off-peak hours, which may > or may not be related to sleeping hours. If peer-to-peer applications > restrained their network usage during periods of peak network usage so > it didn't result in complaints from other users, it would probably > have a better reputation. I am suggesting that ISP folks should be cooperating with P2P software developers. Typically, the developers have a very vague understanding of how the network is structured and are essentially trying to reverse engineer network capabilities. It should not be too difficult to develop P2P clients that receive topology hints from their local ISPs. If this results in faster or more reliable/predictable downloads, then users will choose to use such a client. > The Internet is good for narrowcasting, but its > still working on mass audience events. Then, perhaps we should not even try to use the Internet for mass audience events. Is there something wrong with the current broadcast model? Did TV replace radio? Did radio replace newspapers? --Michael Dillon
Re: Network end users to pull down 2 gigabytes a day, continuously?
> Why would you want to stream in unicast when there are already > models for effective multicast content delivery (see Michael's > list)? *See point above!* The word "multicast" in the above quote, does not refer to the set of protocols called "IP multicast". Content delivery networks (CDNs) like Akamai are also, inherently, a form of multicasting. So are P2P networks like BitTorrent and EMule. If this sounds odd to you, perhaps you don't really understand the basics of either multicast or P2P. Check out Wikipedia to see what I mean: http://en.wikipedia.org/wiki/Peer-to-peer http://en.wikipedia.org/wiki/Multicast If your data absolutely, positively, must be delivered simultaneously to multiple destinations, i.e. time is of the essence, then I agree that P2P and IP multicast are not comparable. But the context of this discussion is not NYSE market data feeds, but entertainment video. The use-cases for entertainment mean that timing is of little importance. More important are things like consistency and control. --Michael Dillon
Re: Network end users to pull down 2 gigabytes a day, continuously?
> > Note that video caching systems like P2P networks can > > potentially serve video to extremely large numbers of > > users while consuming reasonably low levels of upstream > > bandwidth. > > The total bandwidth used is the same though, no escaping > that, someone pays. This is not true. Increased bandwidth consumption does not necessarily cost money on most ISP infrastructure. At my home I have a fairly typical ISP service using BT's DSL. If I use a P2P network to download files from other BT DSL users, then it doesn't cost me a penny more than the basic DSL service. It also doesn't cost BT any more and it doesn't cost those users any more. The only time that costs increase is when I download data from outside of BT's network because the increased traffic reaquires larger circuits or more circuits, etc. The real problem with P2P networks is that they don't generally make download decisions based on network architecture. This is not inherent in the concept of P2P which means that it can be changed. It is perfectly possible to use existing P2P protocols in a way that is kind to an ISP's costs. > If it was only redistributed locally. Even in that case it's not > helping much as it still consumes the most expensive bandwidth (for UK > ADSL). Transit is way cheaper than BT ADSL wholesale, you're saving > something that's cheap. I have to admit that I have no idea how BT charges ISPs for wholesale ADSL. If there is indeed some kind of metered charging then Internet video will be a big problem for the business model. > Or the caches that are being sold to fudge the protocols to > keep it local but if you're buying them we could have just > as easily done http download and let it be cached by existing > appliances. The difference with P2P is that caching is built-in to the model, therefore 100% of users participate in caching. With HTTP, caches are far from universal, especially to non-business users. --Michael Dillon
Re: Network end users to pull down 2 gigabytes a day, continuously?
> > That might be worse for download operators, because people may > > download > > an hour of video, and only watch 5 minutes :/ > So, from that standpoint, making a video file available for download > is wasting order of 90% of the bandwidth used > to download it. Considering that this is supposed to be a technically oriented list, I am shocked at the level of ignorance of networking technology displayed here. Have folks never heard of content-delivery networks, Akamai, P2P, BitTorrent, EMule? --Michael Dillon
Re: Network end users to pull down 2 gigabytes a day, continuously?
> 2. The question I don't understand is, why stream? In these days, when a > terabyte disk for consumer PCs is about to be introduced, why bother with > streaming? It is so much simpler to download (at faster than real-time rates, > if possible), and play it back. Very good question. The fact is that people have been doing Internet TV without streaming for years now. That's why P2P networks use so much bandwidth. I've used it myself to download Russian TV shows that are not otherwise available here in England. Of course the P2P folks aren't just dumping raw DVB MPEG-2 streams onto the network. They are recompressing them using more advanced codecs so that they do not consume unreasonable amounts of bandwidth. Don't focus on the Venice project. They are just one of many groups trying to figure out how to make TV work on the Internet. Consumer ISPs need to do a better job of communicating to their customers the existence of GB/month bandwidth caps, the reason for the caps, how video over IP creates problems, and how to avoid those problems by using Video services which support high-compression codecs. If it is DVB, MPEG-2 or MPEG-1 then it is BAD. Stay away. Look for DIVX, MP4 etc. Note that video caching systems like P2P networks can potentially serve video to extremely large numbers of users while consuming reasonably low levels of upstream bandwidth. The key is in the caching. One copy of BBC's Jan 8th evening news is downloaded to your local P2P network consuming upstream bandwidth. Then local users use local bandwidth to get copies of that broadcast over the next few days. For this to work, you need P2P software whose algorithms are geared to conserving upstream bandwidth. To date, the software developers do not work in cooperation with ISPs and therefore the P2P software is not as ISP-friendly as it could be. ISPs could change this by contacting P2P developers. One group that is experimenting with better algorithms is http://bittyrant.cs.washington.edu/ --Michael Dillon
Re: Phishing and BGP Blackholing
> (All right then, scroll down for content :-)) It is not necessary to quote an entire message when you are only replying to one specific part of it. > Minority? A mail client has been standard-ish for the last three to > four years of upgrade iterations. There are a LOT of mobiles out > there. Granted not many of them are used for e-mail, but that is a One could say that "not many" is a reasonable definition of a minority. So, yes, a MINORITY of users have need for special message formatting. Why should the other 999 million of us need to change the way we do things? > Anyway, I wouldn't write a letter with nothing worth reading on the > first page. I don't write articles with nothing in the first > paragraph. Nor do I, but there is a well-established tradition in written English of the preamble. One could say that a brief quote to set the the context of a statement is perfectly good practice. Of course some people take it to excess like the ones who wrote this declaration a couple of hundred or so years ago: We, therefore, the Representatives of the United States of America, in General Congress, Assembled, appealing to the Supreme Judge of the world for the rectitude of our intentions, do, in the Name, and by Authority of the good People of these Colonies, solemnly publish and declare, That these United Colonies are, and of Right ought to be Free and Independent States, that they are Absolved from all Allegiance to the British Crown, and that all political connection between them and the State of Great Britain, is and ought to be totally dissolved; and that as Free and Independent States, they have full Power to levy War, conclude Peace contract Alliances, establish Commerce, and to do all other Acts and Things which Independent States may of right do. --Michael Dillon
Re: Phishing and BGP Blackholing
> For those of us who read nanog from a mobile device, it's incredibly > annoying to have no content in the first few bytes - a lot of mobile > e-mail clients (all MS Windows Mobile 5 devices and every Blackberry > I've seen) pull the first 0.5KB of each message, i.e. the header, > subject line and the first few lines of text, so the user can decide > which ones are worth reading in full. Why should all 1 billion Internet users change their behavior just because your minority mail-reading system is broken? Hint: Procmail is your friend. Set up your own mail server and run procmail against all incoming email with newline-greaterthan in the first 500 bytes. You can preprocess these messages to do something like strip headers that you don't read and copy the first few reply lines to be first in the message. That way your mobile device will get more bang for the buck than most other people's. Paul Vixie's colo registry may be of help if you need to find a place to stick your own mail server http://www.vix.com/personalcolo/ --Michael Dillon
Re: NATting a whole country?
> > all of Qatar appears on the net as a single IP address. > > I wonder what they use the other 241663 addresses for. Same as you. To address the many machines and networks in Qatar. The existence of a NAT gateway to one portion of the Internet does not remove the need for registered IP addresses. They are still needed to avoid addressing conflicts in the portion of the Internet which is not behind the gateway. --Michael Dillon
Re: Security of National Infrastructure
>> Why is it that every company out there allows connections through their >> firewalls to their web and mail infrastructure from countries that they >> don't even do business in. Shouldn't it be our default to only allow US >> based IP addresses and then allow others as needed? The only case I can >> think of would be traveling folks that need to VPN or something, which >> could be permitted in the Firewall, but WHY WIDE OPEN ACCESS? We still >> seem to be in the wild west, but no-one has the [EMAIL PROTECTED] to be >> braven and >> block the unnecessary access. > > Please don't feed the troll... All those meandering replies full of jokes, puns, political comments and smart remarks do feed the trolls. But a straightforward answer is not troll feeding. The fact is that all those companies out there are PUBLISHING information on their web servers. In order to PUBLISH you must open access to arbitrary members of the PUBLIC. These companies also publish email addresses and invite people to send them email. In order for this email to get through they have to open their incoming mail servers to anyone. This does not mean that their mail infrastructure or web infrastructure is wide open. In most cases only an HTTP load balancer and an incoming-only SMTP server will be accessible directly. If anyone knows of a significant number of companies where this is not the case then I think you have found a potential market for some consultancy services. Rather than whining on NANOG, it would be more productive to find a salesperson to help you get your foot in the door and fix the problems. --Michael Dillon
Re: DNS - connection limit (without any extra hardware)
> What is this group's name? Oh yeah. So that means you have one of > two choices ;-) Smart NANOGers have taken the time to read the NANOG charter here: http://www.nanog.org/charter.html which says... The purpose of NANOG is to provide forums in the North American region for education and the sharing of knowledge for the Internet operations community. --Michael Dillon
Re: Curious question on hop identity...
> Besides, why do you believe the text in an in-addr.arpa record? Or why do > you think the absence of an in-addr.arpa record is meaningful? Back in the old days, say 10 years ago, you could run a network by the seat of your pants using rules of thumb about interpretation of in-addr.arpa records. And you could be quite successful at running a network using such techniques because everybody else was doing pretty much the same thing. Because of this uniformity, you could make a lot of intelligent guesses and resolve problems. However, I think times have changed, there is no longer uniformity among the people making technical decisions about Internet networks and many rules of thumb don't work any more even though they are still out there in network operator folklore. In fact, most people making network architectural decisions about Internet networks don't participate in NANOG any more. Most people making network operational decisions also do not participate in NANOG anymore. It's not just that many people have left NANOG behind, but a lot of newcomers to the industry over the past few years have not joined NANOG because they don't get why it is relevant to them. Not that I'm complaining about the message quoted above. It is a great example of the useful information that one can find in this mailing list. I wish there were more messages like this one, i.e. people sharing info rather than complaints and pleas for help. --Michael Dillon
Re: Clueful Comcast.net Contact Needed
> Can someone clueful from comcast.net contact me offlist please? > Getting through the outer defenses is proving difficult. :-( Do we need some kind of tutorial on how to get through outer defenses and make contact with clueful NOC personnel? The Rockford Files is available on DVD now if you want some general tips... --Michael Dillon
RE: Bogon Filter - Please check for 77/8 78/8 79/8
> B) Threaten the bogon list operator with a lawsuit for falsely claiming your > addresses are bogons and hope they take the simplest path and fix their > list. > > This is a pretty classic case of someone inducing other people to rely on > the accuracy of their data and then offering incorrect data (not arguably > incorrect, manifestly incorrect and most likely negligently so) which those > other people then rely on. It's not just incorrect data. The design of the system used by completewhois is flawed at the core. They only know that certain address ranges are "bogons" at a certain point in time. If their system only reported this fact along with the date for which it is known to be valid, then they would likely win any lawsuits for incorrect data. The fact is, that you can only know that an address range is a bogon at the point in time which you check it and that it WAS a bogon for some past period. For most bogons, it is not possible to predict the future time period during which it will remain a bogon. Any protocol which does not allow the address range to be presented along with the LAST TIME IT WAS CHECKED is simply not suitable for presenting a bogon list. BGP simply is not suitable for this. HTTP/REST, XML-RPC or LDAP could be used to make a suitable protocol. But even better would be to not have any bogons at all. If IANA and the RIRs would step up to the plate and provide an authoritative data source identifying which address ranges have been issued for use on the Internet then bogon lists would not be needed at all. And if people plug their systems into the RIR data feed, then there would be fewer issues when the RIRs start issuing addresses from a new block. IANA would be the authoritative source for stuff like RFC 1918 address ranges and other non-RIR ranges. One wonders whether it might not be more effective in the long run to sue ICANN/IANA rather than suing completewhois.com. --Michael Dillon P.S. As any lawyer will tell you, it is a good idea to make some attempt at solving your issue outside of the courts. Anyone contemplating a lawsuit against ICANN should probably try emailing them and writing a few letters first. Since they are a somewhat democratic structure, it may be possible to get this fixed without lawsuits.
Re: U.S./Europe connectivity
> "You cannae break the laws of physics, Captain!" > > Seriously, LINX is the obvious first step. To find a low latency connection from Chicago to Europe? Somehow I think that he should be shopping locally but it might be useful to use the LINX looking-glass to validate what his local vendors tell him about round trip times. Or he could use a looking-glass in Chicago to measure traffic to various European destinations. LINX, London http://www.linx.net/www_public/our_network/network_tools Equinix, Chicago http://lg.broadwing.net/looking/ If I were in his position I would make the rounds of all vendors in Chicago, ask for prices and latency data, then check their latency numbers using various looking-glass sites. If a vendor gives out numbers that vary significantly from what you can measure then I would want a detailed explanation of why that is. --Michael Dillon
Re: U.S./Europe connectivity
> I am doing some work on a network in central Illinois that is > currently peering with Sprint and McLeod. They have a number of > customers in the U.K. and they want to reduce latency to that part of > the world. Make sure they're not trying to reduce latency below the speed of light in fibre. Make sure that your client understands that they will never achieve the same latencies trans-Atlantic as they achieve within the state. We recently had to haul back one of our over-eager account managers who was trying to sell a low-latency solution that was about 3 times faster than the speed of light in fibre. BTW, the speed of light in fibre is roughly equal to the speed of electrons in copper and roughly equal to two-thirds the speed of light in a vacuum. You just can't move information faster than about 200,000 km/hr. --Michael Dillon
Re: Fwd: The IESG Approved the Expansion of the AS Number Registry
> Thanks very much for this link (and the summary). I see an interesting > (if not surprising) trend in Advertised AS Count. Up until 2001 it was > accelerating... and after 2001 its stayed linear. However, unadvertised > AS count which was basically stagnant has increased markedly before then. Those are not unadvertised ASNs. Those are only ASNs which have been issued but are undetectable by his monitoring tools. That doesn't mean that they are not advertised, just that his tool cannot detect them. Given the fact that the Internet is now thoroughly global with rich interconnectivity in most regions of the globe, it is hardly surprising that lots of ASNs do not get advertised globally. The trend you see is likely cause by rich local interconnectivity becoming the norm rather than a few circuits from the capital city to some big U.S. city. --Michael Dillon
Exotic meeting locations in North America
There really is no need for all NANOG meetings to have the same format. In fact, if we accept the idea of varying formats, then some of the cost issues can be tamed. For instance, one full meeting, one regional meeting, and one special-focus meeting per year. The full meeting could be the one that is done in conjunction with ARIN in a major center with full free networking, beer and gear etc. The regional meeting would be in a smaller city with the expectation that the majority of attendees are from the local area and don't have access to big travel budgets. And the special focus meetings would target some specific topic and pick a location to match. Some of the regional and special focus meetings would not supply comprehensive free Internet access. If Internet access is available people would pay for it and expect bandwidth limitations and higher than normal latency. Depends on the location. Here are some exotic locations that could work with a special focus. Iqaluit, the capital of Nunavut is rather exotic. The native language is neither English nor French nor Spanish. It has the issues of remoteness and reliance on satellite telecommunications. New Orleans has dealt dramatically with disaster recovery and rebuilding infrastructure. It is exotic because it is still in the process of rebuilding unlike most American cities. St. John's, Newfoundland - a British colony until 1949 when it joined Canada, this is located on a large island, has a history in trans-atlantic telecommunications and still has a certain amount of undersea fiber connectivity. Montpelier, Vermont is the smallest state capital in the USA, located in the Vermont,New Hampshire, Maine area which is rather more rural than the average in the USA as well as being somewhat mountainous terrain. If you don't count New Orleans before Katrina, I'd guess that well over 90% of NANOGers have never been to any of these four cities. Other special focus areas might be: Government and the Internet, Government and IPv6 - Washington DC. The Research Community and the Internet - Ann Arbor MI Network Security from a Military Viewpoint - Sierra Vista AZ near US Army's CECOM-ISEC headquarters Strategic Aspects of Network Security - Harrisburg PA not far from US Army War College Strategic Studies Institute in nearby Carlisle The idea of regional meetings is mainly to have a scaled down NANOG to reach a much wider audience that does not have a large conference travel budget. This is rather similar to RIPE's meetings in Qatar, Moscow, Bahrain, Nairobi and Tallinn. The idea of special focus meetings is to do something entirely new, perhaps redefining the NANOG role and audience in the process. It is clear that the traditional NANOG audience is shrinking because the traditional Internet provider has been mostly replaced by larger general telecommunications providers. The same old topics and same old restricted set of participants doesn't have enough future potential to keep NANOG running in the long term. Special focus meetings can help bring in new blood. ------- Michael Dillon Capacity Management, 66 Prescot St., London, E1 8HG, UK Mobile: +44 7900 823 672Internet: [EMAIL PROTECTED] Phone: +44 20 7650 9493Fax: +44 20 7650 9030 http://www.btradianz.com One Community One Connection One Focus
RE: IP adresss management verification
> > They ask for it. > Is the policy still that dedicated IP addresses for web > hosts *should* only be used when technical justification > exists? I really wish it would change to a requirement > as we very frequently get new hosting customers who get > angry when they find their site that doesn't have SSL > or any other technical reason for a dedicated IP ends up > on a shared IP when their old host didn't do that or > would sell them a dedicated IP for $5/month, etc. SSL is a technical justification for separate IP addresses for web hosts. Virtual servers is another technical justification for assigning multiple IP addresses to a single physical server. ISPs should really make it clear to their customers which features are included in a product and which ones are not. There is no reason why you could not have offered a "web hosting" product with no SSL capability at one price point and a "secure web hosting" product with SSL at another pricepoint. What is the point in leaving things vague? And if you look at this template http://www.arin.net/registration/templates/net-isp.txt updated in September, it no longer mentions web hosting. --Michael Dillon
Re: IP adresss management verification
>I'm curious on how regional RIR which allocates ip address, verifies the > usage pattern info provided by their members in their application process. It's quite simple, really. They ask for it. If the iformation that you provided with your application does not answer their questions, they ask you for more information. I assume that all the RIRs will sign an NDA with you, certainly ARIN does this. ARIN may also ask for corporate confidential information in order to verify your application so they have strict internal security policies to keep it confidential. Some people send detailed network diagrams, purchase orders for routers/switches/circuits, sales history data with projected trends, customer lists, etc. If you need specific details, just ask your RIR. --Michael Dillon
Re: [c-nsp] [Re: huge amount of weird traffic on poin-to-point ethernet link]
> how about PORN-SOURCE, COMMUNIST-SOURCE, DEMOCRACY-SOURCE, > TERRORIST-SOURCE, RIGHT-WING-CHRISTIAN-SOURCE, COURT-ISSUED-LIBEL-CASE-SOURCE > > be careful before you open such a pandoras box... The box was opened a long time ago. In an Internet context, there are many email blacklists which apply various different criteria for inclusion, therefore, they are essentially publishing different attributes. In a social context, freedom of religion is a long-accepted principle and various religions publish lists of literature that is either acceptable or unacceptable. If a network operator finds a business case for supplying service only to right wing organizations and blocking network traffic from communist sources then what is wrong with that? The principle of the Internet is that network operators run private networks and set their own policies independent of regulators and governments. > will this scale? The fact that the database has multiple attributes to assign to address ranges makes it more likely to scale. > who will want to use it? People who find some value in dynamically filtering Internet traffic based on a trusted source for filters. > can it be exploited? Virtually anything can be exploited. Smart network operators do not hardwire their routers to a 3rd-party BGP feed. Instead they pull that feed into their operational support systems where it can raise alarms so that a human being can decide whether to stop or start filtering a particular range. Or else they make some kind of 2-party binding contract with SLAs and penalties such as a transit contract or a peering agreement. > what sort of liability do you take on by becoming responsible for > policing the Internet? Who said anything about policing the Internet? This is all about identifying address ranges who source various kinds of traffic that some network operators do not wish to transit their networks. Every network operator has an AUP for their own customers and peers. This merely extends that to 3rd parties who wish to transit the network. --Michael Dillon
Re: [c-nsp] [Re: huge amount of weird traffic on poin-to-point ethernet link]
> > If there were some way to have a feed of real bogons, > > i.e. address prefixes that are *KNOWN* to be bogus at > > the point in time they are in the feed, that would be > > useful for filtering. And it would likely be a best practice > > to use such a feed. > > > > But at the present time, such a feed does not exist. > > http://www.cymru.com/BGP/bogon-rs.html That is not a feed of routes that are known to be bogus. That is a feed of routes that use addresses which have not been allocated by IANA to an RIR. There are many bogus routes that are not included in the Cymru feed. For instance, RIR address ranges that have not yet been allocated ISP address ranges that have not yet been assigned Assigned address ranges that are not announced by the assignee. Address ranges from which a high percentage of the traffic is SPAM, i.e. a network owned by spammers. I am arguing that it is better to start with a database that allows several attributes, both negative and positive, to be associated with address ranges. Then build a feed from that, in fact, allow the user to specify which attributes they want in their feed. One size fits all just doesn't work. --Michael Dillon
Re: [c-nsp] [Re: huge amount of weird traffic on poin-to-point ethernet link]
> WRT acls, I would suggest any acl is a bad idea and only a dynamic > system such as rpf should be used, this is because manual filters > that deny bogons has the same issue as BGP filtering in that it can > go stale and you drop newly allocated space. Your comment implies that ACLs are static and must be configured manually. In this day and age of automated systems, that is no longer true. Anyone who wants to can easily implement dynamic ACLs. They will be slightly less dynamic than a routing protocol, but ACLs do not have to be manually configured and do not have to be static. Of course, on some hardware ACLs have a significant CPU impact, but that is less of a factor than it used to be. --Michael Dillon
Re: odd hijack
> My question to the community is, > what kind of misconfiguration could cause this set of prefixes to be > announced? > 11.0.0.0/8 > 12.0.0.0/7 > 121.0.0.0/8 > 122.0.0.0/7 > 124.0.0.0/7 > 126.0.0.0/8 > 128.0.0.0/3 etc ... This looks to me like some large multinational leaked their internal announcements to an ISP. It is not unusual for large companies to use random unregistered /8 blocks in their internal networks. There are all kinds of applications that need to talk across networks which do not need any Internet connectivity or any direct connectivity to general use workstations. This network traffic would normally be hidden inside some kind of VPN on the same infrastructure as other corporate traffic. So to answer your question, first look for all the ways that a misconfiguration could allow routing information to leak out of some flavor of VPN. --Michael Dillon
Re: [c-nsp] [Re: huge amount of weird traffic on poin-to-point ethernet link]
> > The craziest stuff that gets announced isnt in the > > reserved/unallocated realm anyway so the effort seems to be > > disproportional to the benefits... and most issues I read about with > > reserved space is packets coming FROM them not TO them > > Steve's 100% spot-on here. I don't have bogon filters at all and it > hasn't hurt me in the least. I think the notion that this is somehow > a good practice needs to be quashed. I think there is a terminology problem here. People think that "bogons" means "bogus routes". From that they infer that bogus routes should be filtered and use the Cymru feed because it seems to be a no-brainer. The problem arises because the Cymru feed only contains the low-hanging fruit. It only refers to address ranges that *might* be bogus and which are easy to identify. The problem is that if you pick this fruit, it soon goes rotten and you end up filtering address ranges which are in use and almost certainly not bogus. If there were some way to have a feed of real bogons, i.e. address prefixes that are *KNOWN* to be bogus at the point in time they are in the feed, that would be useful for filtering. And it would likely be a best practice to use such a feed. But at the present time, such a feed does not exist. Also, I think that anyone contemplating creating a new feed should give some thought to what they are doing. It would be very useful to have a feed or database which can assign various attributes to address ranges. When there is only one possible attribute, bogon, then the meaning of the attribute gets stretched and the feed becomes useless. But if there are many attributes such as UNALLOCATED, UNASSIGNED, DOS-SOURCE, SPAM-SOURCE, RIR-REGISTERED then it starts to look interesting. Some networks might like to filter based on several attributes, others will just filter those with the DOS-SOURCE attribute. Obviously, it would require lots of cooperation for some of these such as UNASSIGNED, but perhaps the Internet needs to move towards more cooperation between network operators. --Michael Dillon
Re: Urgent need for bandwidth in Chiswick/London
> Does anyone know of any MAN (or anything else for that matter) options > at this location? Yes, somebody does know but they are unlikely to be on this list. In general, for any commercial building in any city, there is a building manager who takes care of utilities, air conditioning, heating systems, etc. This person or persons will be aware of any and all telecommunications circuits into the building and which companies send technicians to monkey around with the comms closet. That should be your first port of call. If that person is too hard to get answers from, next best is to talk to neighbouring commercial buildings because MANs are built in rings so chances are that one of the companies connected to neighbouring buildings will also be in yours. Third port of call is to contact the sales departments of all the companies offering MAN services in your city. Since they have a chance to win some business, they will be happy to research their internal records to find out whether or not they have infrastructure in your building. Fourth port of call is your other technical contacts in the city/country in question. And maybe in the fifth position, or lower, is a general mailing list like NANOG. My question is, did you actually go through all the above BEFORE posting to NANOG? --Michael Dillon
Re: BCP38 thread 93,871,738,435 + SPF
> > > How is this attack avoided? > > > > Sounds like the attack is inherent in SPF. In that case, > > how did the thread about dns providers and rfc compliance morph into SPF > and spam discussions? Ask Doug Otis. He stated that SPF sets the stage for DDoS attacks against DNS servers. Presumably he said this because it points to another *COST* of DDoS that could be used as a business justification to implement BCP38. Or you could look at it as a weakness of SPF that should be used as a justification for discouraging its use. After all if we discourage botnets because they are DDoS enablers, shouldn't we discourage other DDoS enablers like SPF? --Michael Dillon
Re: Extreme Slowness
> Which begs the same question I've asked in the recent past: then > what *is* a good diagnostic tool? If ICMP "is not the best way to > test", then what is? What other globally-implemented layer 3 or > below protocols do we have available for troubleshooting? > > Sure, UDP-based traceroute still relies on ICMP TTL exceeded > responses to work. I've no idea what TCP traceroute relies on, > as I haven't looked at it. I love it when people answer their own questions and tell us that they are lazy, to boot. For the record, TCP traceroute and similar TCP based tools rely on the fact that if you send a TCP SYN packet to a host it will respond with either a TCP RST (if the port is NOT listening) or a TCP SYN/ACK. The round trip time of this provides useful information which is unaffected by any ICMP chicanery on the part of routers or firewalls. A polite application such as TCP traceroute will reply to the SYN/ACK with an RST packet so it is reasonably safe to use this tool with live services. Of course, even TCP packets can be blocked or dropped for various reasons so this is not a 100% solution. However, if you want to avoid ICMP filtering or low precedence, then TCP traceroute will help. --Michael Dillon
Re: BCP38 thread 93,871,738,435 + SPF
> How is this attack avoided? Sounds like the attack is inherent in SPF. In that case, avoiding it is simple. Discourage the use of SPF, perhaps by putting any SPF using domain into a blacklist. Eventually, people will stop using SPF and the attack vector goes away. --Michael Dillon
Re: register.com down sev0?
> but i am not foolish enough to believe > that religious ranting on mailing lists is gonna change anyone from > doing what makes business sense for their network. Indeed! And it is not going to change the minds of the majority of network operations folks who are not on the NANOG list nor the majority of telecoms executives who are also not on the NANOG list. Back in the old days, the NANOG list did hold the majority of Internet operations folks so new ideas like flap dampening were able to spread quickly. But those days are long gone. NANOG still has an important educational role but it is no longer based on being part of the old boys club and knowing the secret handshake. In other words, there is no cohesive society of network operators which can be swayed by attempts at social engineering like shaming or cajoling. BCP 38 has had its day. Nowadays, it is more important to look at how to mitigate current DDoS techniques and to describe the larger problem and look for larger solutions. However, any attempt at larger solutions require a large amount of humility because nobody can say for sure, what will work and what won't. The fact remains that there is not a good technical method for mitigating large scale distributed DDoS that results in LARGE TRAFFIC FLOWS ENTERING A NETWORK FROM ALL PEERED ASES SIMULTANEOUSLY. Perhaps if we could find a way to allow the attacked AS to set ACLs automatically in all the source AS networks, that would help mitigate these attacks. For instance, consider a set of ASes which all install an ACL-setter box. These boxes all trust each other to send-receive ACL setting requests through a trusted channel. The owner of a box sets some limits on the ACLs that can be set, for instance n ACLs per AS, max ACL lifetime, etc. And the box owner also decides the subset of their routers which will accept an ACL for a given address range. Then when an attack comes in, the victim AS uses some tool to identify large sources, i.e. a CIDR block that covers some significant percentage of the source addresses in one AS. They then issue an ACL request to that AS to block the flow and the ACL takes effect almost instantaneously with no human intervention. Yes, this can result in some IP addresses being blocked unfairly, but the DDoS traffic levels often have the same impact. In any case, the AS holding the destination address is the one doing the blocking even though the mechanism is an ACL inside the source AS network. On the technical side, it is not a complex problem to put such a system in place. The complexity is largely in getting network operators to come to an agreement on the terms under which operator A will allow operator B to set ACLs in operator A's network. Until network operators see DDoS as a significant business problem, this will not happen. Note that a "business problem" does not refer solely to the direct costs of mitigating a DDoS attack. It also includes the indirect fallout which is harder to measure such as loss of goodwill, missed opportunities, etc. --Michael Dillon
re: passports for NANOG-39, Toronto
> http://travel.state.gov/travel/tips/regional/regional_1170.html > December 31, 2007 - Passport required for all land border crossings, as > well as air and sea travel. If someone wants to go but does not have a passport for whatever reason, i.e. last minute travel plans, then it is possible to fly to Buffalo NY and make a land crossing from there, i.e. bus or rental car. If you do want to take a rental car across the border, you have to notify your rental company so they can issue a non-resident insurance card for you. As long as you have a US driver's licence this is fairly routine. Cross the bridge to Canada and take the QEW all the way to Toronto. http://www.buffaloairport.com/ You could do the same fly-drive via Detroit but there is a lot more driving. --Michael Dillon P.S. Now that you have your shiny new passports, don't just stop at Canada. There's a whole world out there.
RE: Collocation Access
> I'm not exactly sure why these sites want to retain ID, but I think it > goes along with the big weight that is connected to the gas station bathroom > key. They want to make sure you return your cabinet keys (if any), > temporary pass (if any), etc. Legal risk or not, can you think of a better > way to get someone to return to the security desk to sign out? Until then, > these sites will continue this practice. a) cash deposit b) heavy weight attached to cabinet keys and temporary pass c) bulky object attached to cabinet keys and temporary pass In high school, our "data centre" keys were attached to a few links of chain bolted onto a chunk of 2 x 4. I never mislaid them. I remember at least one place where I received a plastic card key similarily attached to a few links of chain welded to an broken wrench. Why couldn't ID cards be treated the same way? For that matter, in these days of RFID badges, why can't colo centers issue "magic wands", 3 foot long rods tipped with an embedded RFID tag? They would not fit in pockets or briefcases etc. They would function identically to the RFID tags embedded in credit-card sized plastic but they would never "get lost". Perhaps what we have here is another "failure of imagination" like the one cited in the 9/11 report. --Michael Dillon
Re: Broadband ISPs taxed for "generating light energy"
> A Cisco ZX GBIC produces a max of 4.77 dBm (or less than 4mw). 4mw > corresponds to 35 watt hours in one year. .035 kwh per year costs 34.5 cents per year using the average US electricity cost in March 2006 of 9.86 cents/kwh. Since the energy flow could be bidirectional, one of the two parties receives a net benefit of up to 34.5 cents. If a broadband provider offers customers a free gift such as a hat, does this make them into a hat retailer for tax purposes? --Michael Dillon
Re: Broadband ISPs taxed for "generating light energy"
> "In the process of > data transmission, other than light energy, no other elements are involved and > the customers are paying for the same. This proves that light energy > constitutes goods, which is liable for levy of tax. Therefore, the State has > every legal competence and jurisdiction to tax it," the department has > contended. Sounds reasonable to me. Since the sale of energy is usually measured in kilowatt-hours, how many kwh of energy is transmitted across the average optical fibre before it reaches the powereda mplifier in the destination switch/router? I'd like to see some hard numbers on this. The light shining down optical fibres is laser light. There exist medical devices which are powered by laser light shining through the tissues. There are also some types of satellite devices which can receive power from ground-based laser beams. The crux of this issue is the actual measurement of power transmitted which will turn out to be very small. --Michael Dillon
Re: that 4byte ASN you were considering...
> Well, it will break an applications that considers everything > consisting of numbers and dots to be an IP address/netmask/inverse > mask. I don't think many applications do this, as they will then > treat the typo "193.0.1." as an IP address. An application using "[0123456789.]*" will not break when it sees the above typo. 193.0.1. *IS* an IP address-like object and any existing code will likely report it as mistyped IP address or mask. > It won't break applications > that check if there are exactly 4 numbers in the 0-255 range and 3 dots. True, however my point is that I do not believe that all existing applications do this. Therefore, changing their input in an unexpected way will break them. > >The real question is what does the notation 1.0 add that the > >notation 65536 does not provide? > > It is (for me, and I guess most other humans) much easier to read and > remember, just as 193.0.1.49 is easier to read and remember than > 3238002993. It also reflects that on the wire there are two 16 > bit numbers, rather than 1 32-bit number. In my experience, ISPs do not transmit numbers by phone calls and paper documents. They use emails and web pages which allow cut'n'paste to avoid all transcription errors. And I know of no earthly reason why a general written representation needs to represent the format of bits on the wire. How many people know or care whether their computer is bid-endian or little endian? > 1. If you are a 16-bit AS speaker (ASN16), then AS65536 is not just > the next one in the line, it is an AS that will have to be treated > differently. The code has to recognize it and replace it by the > transistion mechanism AS. And how is a special notation superior to if asnum > 65535 then process_big_as else process_little_as In any case, people wishing to treat big asnums differently will need to write new code so the dot notation provides them zero benefit. > 2. Just as people having used the regexps that you mentioned, I'm > also certain that people have used unsigned short int's or > signed long int's in their code. Typically ISPs are using apps written in higher level languages which are more likely to treat integers as 32-bit signed quantities. In any case, this is a length issue, not an issue of notation. > In short, like it or not, you will have to check and update your tools > anyway. My point is that if we do NOT introduce a special notation for ASnums greater than 65536, then tools only need to be checked, not updated. If your tool was written by someone who left the company 7 years ago then you might want to do such checking by simply testing it with large as numbers, not by inspecting the code. The dot notation requires that somebody goes in and updates/fixes all these old tools. --Michael Dillon
Re: that 4byte ASN you were considering...
> > - 'Canonical representation of 4-byte AS numbers ' > > http://www.ietf.org/internet-drafts/draft-michaelson-4byte-as-representation-01.txt > > and what is good or bad about this representation? seems simple to me. > and having one notation seems reasonable. what am i missing? It breaks any applications which recognize IP address-like objects by seeing a dot in an otherwise numeric token. For the purposes of parsing a string into internal representation, an application can treat IP addresses, netmasks and inverse masks identically. We all know that the Internet is awash in homegrown scripts written in PERL or TCL or bash or Ruby or Python. It is likely that many authors have, in the past 15 years, written scripts which contain regular expressions like "[0123456789.]*" to match a string containing only digits and the period. Those scripts will be confused by this AS number notation. Also, any script which "recognizes" IP address-like objects when it hits the first period in a numeric string. The real question is what does the notation 1.0 add that the notation 65536 does not provide? All I can see is that it adds the risk of broken scripts and the confusion of AS numbers that look like decimal numbers. If the IETF had really wanted to create a universal notation then they should have recommended that AS numbers be represented in the form AS65536 which is completely unambiguous. When IP addresses were created, it was important to indicate the boundaries between the network number and the host address. Originally, the periods represented this boundary for the three classes of IP address, class A, class B and class C. Long ago, we removed this classfulness attribute, but the notation remains because lots of applications expect this notation. So why on earth are we changing AS number notation today? --Michael Dillon
Re: Armed Forces Information Service.
> Could someone responsible for the armed forces information service > please contact me off list. Thanks. Which AS are you refering to? Why didn't you mention the AS in your posting to the list? --Michael Dillon
Re: Anyone from state of illinois or ATT on list?
> One of the state of illinois' servers is attempting to breach > security in our network and their arin abuse contact doesnt exist. Then contact their upstream network provider. If someone doesn't publish an abuse contact in the normal way, it generally means that there is nobody competent available to be contacted. You don't really know that it is a state of Illinois server. All you can tell from network diagnostic tools is that the traffic appears to originate from ISP X who provides Internet access to the state of Illinois. Make them responsible for their traffic. --Michael Dillon
Re: icmp rpf
> The non-announcers, because they're also breaking PMTUD. If you're not sure what benefits PMTUD gives, you might want to review this page: http://www.psc.edu/~mathis/MTU/index.html --Michael Dillon
Re: Topicality perceptions
> One of the biggest issues with the list as I've seen from time to > time from my perspective, is the definition of "operations". So on a > quick breakdown of the logical definition of NANOG, I derive > "Operations of the North American Network". The problem with this > stems from far too many bastardizing their own definition of what it > should be. Please don't contribute to the bastardization. Section 3 of the NANOG charter states: The purpose of NANOG is to provide forums in the North American region for education and the sharing of knowledge for the Internet operations community. You can read the full charter here: http://www.nanog.org/charter.html By your definition, Cat's recent request for outage information about Telehouse North would be off-topic. But according to the NANOG FAQ here: http://www.nanog.org/listfaq.html outages are on topic. Obviously, network infrastructure tends to span political borders and geographic borders, therefore it is not unusual that Cat has an infrastructure issue in Europe to deal with. On your first point, the fuzziness and lack of clarity of what network operations issues belong on this list, I agree. The FAQ is never posted on the list so it has become an obscure document hidden away on a little-used website. It needs to be promoted more and I think it needs to be updated to communicate more clearly. > These are off-topic but I wouldn't trade em for the world. > I've learned much from them, as have I from all sorts of posts on > topic or not. I agree with you. Unfortunately some old-timers would rather see a return to the old days when network ops and engineering was an obscure passtime only understood by those who knew the secret handshake and were admitted to the inner circle. They forget that NANOG's major role has been in educating the new people who have flooded into the net ops community as the Internet grew and grew and grew. --Michael Dillon
Re: tech support being flooded due to IE 0day
> To the people who say we throw in the towel and just say "Gadi will never > stop posting off-topic crap, so why bother trying to correct him?", I'd > suggest that this is a self-defeating attitude. Not only because Gadi > could actually be posting useful stuff if set on the right path as to what > is appropriate and what is not, but because 10,000 other people are going > to be reading that post and thinking that this is appropriate subject > matter. One off-topic post you can delete, but an entire list which has > been co-opted by off-topic material can not be fixed. I agree with you 100%. Please give us your list of *ALL* the topics that you think are appropriate for this list. --Michael Dillon P.S. Note that I do not agree that anyone has yet tried to "correct" Gadi. All I have seen is bellyaching on a personal level, i.e. person A does not like person B's message. To set everyone on the right path we need a description of the path itself.
Have you really got clue?
> that, and a thread where half of the posts are from the > initial poster himself anyway. but then, happily watching > him, at least he is creative in topics... i am mentally > killfilling his threads anyway, less and less relevant. > it is scary what stuff is discussed lately. > > -ako OK, Alexander Koch. You apparently have clue and you apparently know what *IS* on topic for this mailing list. Instead of posting an off-topic message like the one above, kindly post a message listing *ALL* of the topics that belong on this list. And if anyone else here thinks they know what is on topic, please tell us. I am getting bored by the flood of negative messages that say only "You can't say that here". Please stop telling us what you cannot say on NANOG. If you really must register your discontent with a message, then at least take the time to list some of the topics that belong on the list. What is NANOG all about? What is relevant to network operations? Is NANOG a narrowly focused technical list for a small group of technical specialists? Or is it some kind of broader industry-focused list that covers many issues relevant to the industry? --Michael Dillon
Re: tech support being flooded due to IE 0day
> i've > assumed that the hardcore bgp engineering community now meets elsewhere. Or perhaps BGP engineering hasn't changed in so many years that it is now more than adequately covered by books, certificate courses, and internal sharing of expertise. Lists are good for things that are new or confusing or difficult. BGP no longer fits into those categories. > (c) the flames completely outweigh gadi's own original posts, Words of wisdom. I was wondering when someone would point this out. > and (d) some of > the folks lurking here actually tell me that they benefit from gadi's stuff. And, no doubt, they tell Gadi too which is why he continues to post on this list and does not seem to be wounded by the flaming arrows sent his way. > ISC Training! October 16-20, 2006, in the San Francisco Bay Area, > covering topics from DNS to DHCP. Email [EMAIL PROTECTED] Now that is on topic. Maybe we need more advertising on the list to make people happy? --Michael Dillon
Re: IPv6 PI block is announced - update your filters 2620:0000::/23
> Yes, please, let's have that flamewar all over again... Or you couldjust read > one or more of the previous flamewars and spare us another round. Here's a > starting point: The problem with this suggestion is that it doesn't have an end-point. If someone would summarize both the pros and the cons of bogon filtering on a page at http://nanog.cluepon.net then it would be reasonable to say that the poster should go elsewhere for their information. Until people actually populate that wiki with useful information, we just have to accept that things will be rehashed and rehashed on this list. --Michael Dillon
Cyber Storm Findings
A quote from the DHS's recently released report about their Cyberstorm exercise in Feb: http://www.dhs.gov/interweb/assetlibrary/prep_cyberstormreport_sep06.pdf Finding 3: Correlation of Multiple Incidents between Public and Private Sectors. Correlation of multiple incidents across multiple infrastructures and between the public and private sectors remains a major challenge. The cyber incident response community was generally effective in addressing single threats/attacks, and to some extent multiple threats/attack. However, most incidents were treated as individual and discrete events. Players were challenged when attempting to develop an integrated situational awareness picture and cohesive impact assessment across sectors and attack vectors. And a question: Do network operators have something to learn from these DHS activities or do we have best practices that the DHS should be copying? --Michael Dillon
Re: renumbering & IPv6
> The 8xx system is the one which maps to domain names, > not the standard land-line system. In the United States, due to number portability regulations, the standard land-line phone numbers also map to domain names because they are no longer used for routing calls. In the UK, mobile phone numbers also map to domain names because of regulations that allow you to switch mobile network operators and maintain your phone number. > Perhaps a customer who wanted to make IP addresses > "portable" would pay a fee to the ISP whose addresses > they are, and maintain redirection equipment to the > "real" IPs... And perhaps the price of doing so would > actually be higher than just keeping a T1 to that > first provider... There are people who are proposing a mechanism like that in order to do a new type of multihoming in IPv6. http://www.ietf.org/html.charters/multi6-charter.html --Michael Dillon
RE: Watch your replies (was Kremen....)
> > Perhaps the list should be turned into a wiki; > I might just to watch the hilarity. Is there any real interest in this? Do we want another wiki to compete with http://nanog.cluepon.net ? Mediawiki is a good idea, but proliferation is not so good. Also, if you want to contribute, why not write up a page or two for the existing wiki? --Michael Dillon
Watch your replies (was Kremen....)
> It's insulting > when you trim the message to a shorter statement that you are > responding to. The other 18 lines may not have been important to this > particular response but they were not content free. If your content was in any way, interesting, then people will have read it in the message that you posted. I see no need to repeat a bunch of irrelevant text when I am only replying to one point in your email. Personally, I wish more people would trim away all the irrelevant junk when replying. On the other hand, in the corporate world I find that the habit of top posting is very useful to me. I often see things that were never intended to be sent to me and I often discover that the previous replies in a thread betray the fact that the writer did not read or did not understand the original message. But on a mailing list, trimmed replies are superior. --Michael Dillon P.S. are the standards of this list so unclear that Darcy and I have to discuss this? Who is right?
Re: Kremen's Buddy?
> Ever notice the only folks happy with the status quo are the few who have > already have an intimate knowledge of the ARIN allocation process, and/or > have the right political connections to resolve the "issues" that come up > when dealing with them? In this case, the "right political connections" means that we go to the ARIN public policy meetings, participate on the ARIN Public Policy mailing list http://lists.arin.net/mailman/listinfo/ppml and work to get changes to ARIN policies approved. Like any other kind of politics, you lose some and you win some because, in the end, ARIN doesn't do things my way or your way. That's the politics of compromise. > Try looking at it from an outsider's point of view instead. If you're new > to dealing with ARIN, it is not uncommon to find the process is absolutely > baffling, frustrating, slow, expensive, and requiring intrusive disclosure > just shy of an anal cavity probe. Sounds like my very first time trying to get an IP network functioning between an SCO Xenix server and a bunch of DOS and Windows 3.0 workstations. Education and experience do work wonders to solve this problem. --Michael Dillon
RE: Kremen's Buddy?
> "From each according to his abilities, to each according to his needs" could > be replaced with "From each according to the ARIN fee schedule, to each > according to our impossible to decipher allocation templates". Marx would be > proud! Centrally managed economic systems seem so wonderful on paper - > that's why so many otherwise very smart people have championed the idea. > Real world experience, on the other hand, has shown that capitalism is the > worst possible method for distributing resources - except for all the other > methods, which are even worse. Your tirade is a typical example of Marxist rhetoric. It may sound good on paper but it does not work in the real world. For an authentic example of such intellectual argument, read this http://www.marxists.org/archive/marx/works/1875/gotha/ch01.htm by a famous German intellectual whose tomb is in London's Highgate cemetery. It's only the 1st of 4 chapters, but if you stick to it, not only will you begin to understand why theoretical arguments can be pointless, you will also discover the origin of one of the most well known quotes in the sphere of economics. > And, if people have noticed, the Internet is what we use to make money, > these days - at least, the folks on this list. The fact that the Internet is what we use to make money is the main reason why most of us do not want to make any major changes to the infrastructure that enables the Internet such as ARIN itself. If it ain't broke, don't fix it. > My opinion is that ARIN should use some of its not inconsiderable warchest > and hire some economists to do some real work on modalities for address > distribution (i.e. give some grants). Aside from the practical utility, some > real science around this topic would be of great intellectual benefit. As I said, a meaningless intellectual exercise... --Michael Dillon
Re: Kremen's Buddy?
> The fact that there is a lot of space assigned/allocated and not used > in any easily observable way is well known to those who track the > address exhaustion issue, I think. The fact that addresses are not used in an observable way does not imply that the addresses are not used at all. It simply means that the observation techniques used are not perfect. --Michael Dillon
RE: Kremen's Buddy?
> It seems to me that this nicely illustrates a major problem with the > current system. Here we have large blocks of IP space that, by their > own rules, ARIN should take back. It all sounds nice on paper, but > clearly there is a hole in the system whereby ARIN doesn't know and > apparently has no way of figuring out that the space is no longer in > use. Or maybe it means that ARIN has priorities and recovering this space is low on the priority list. Anyway, you are wrong. ARIN does have a way of figuring out that the space is no longer in use. When some sucker buys the addresses and tries to use them, they will find out that they must first update ARIN's records. And when they do that, ARIN will learn about the deal. At that point, they have to justify their address space just like anyone else, and only get to keep the amount of address space which they can justify. The fact that there are few suckers around to buy these addresses means that these block have been kicking around for a long time. But if there is ever a crunch for IPv4 address space, you can bet that ARIN members will empower ARIN to act unilaterally and take back the space. > but the way things currently work it seems like if you can > justify a block today, it's yours forever even if you stop actively > using it. You haven't read through ARIN's policies yet, have you? --Michael Dillon
Re: Commodity (was RE: [Fwd: Kremen ...])
> Erm, Uranium *is* a commodity. Last week's spot price was > $52 a pound for U3O8. It's a small market in terms of numbers > of players but it's still an open market in the economic sense. > 102 million pounds were traded in 2004. Hedge funds are players > in the uranium market (source: www.uxc.com, home page) I don't know where you got that figure but the website you reference states that in 2005 only 35 million pounds were traded in 107 transaction. I think most people will agree that any item for which only 107 transactions are concluded in a year is not terribly liquid. According to this http://www.cbot.com/cbot/pub/cont_detail/0,3206,1248+21215,00.html in Chicago alone, counting only trades for 100 oz. unit size, there were 15,544 contracts. Add to that the fact that you can buy and sell gold in any major bank in any major city as well as in most large jewellry stores and you have a very liquid commodity indeed. --Michael Dillon
Re: Commodity (was RE: [Fwd: Kremen ...])
> > Since IP addresses are tightly tied to the network > > architecture, how can they ever be liquid? > > How are PI addresses tightly tied to network architecture? What percentage of the total IPv4 address space is PI? If non-PI addresses are not property then how do PI addresses gain that attribute? --Michael Dillon P.S. PI addresses get configured into devices just the same as non-PI addresses. If you could sell a PI block then you would be faced with the prospect of renumbering all those devices. DHCP makes end-user devices pretty easy, but devices in the NETWORK ARCHITECTURE pose more of a problem. In addition there are some people who use IP addresses encoded in hardware in a non-mutable fashion. Those people will apply for PI allocations which, on average, makes PI addresses more tied to the hardware than non-PI. But the important points are not the ones mentioned in this postscript.
Commodity (was RE: [Fwd: Kremen ...])
> You make an incorrect assumption - that IP addresses are currently free > (they are not, in either money or time) and that commoditizing them will > increase their cost (there is significant evidence it will not). You seem to think that commoditizing IP addresses will reduce their cost. Commoditization is changing an illiquid resource into a liquid resource, i.e. one that can readily be converted to cash in an open market. Since IP addresses are tightly tied to the network architecture, how can they ever be liquid? If they cannot become liquid then they can never be a commodity in the first place. For example, let's compare gold and uranium. Both metals are very valuable. Gold can be bought and sold at any time on an open market. It is a commodity. But uranium is not as liquid. There are few buyers and sellers. Trades happen too infrequently to establish an open market. There are restrictions on posession and transport of the material. In the end, uranium is not a commodity and is not liquid. IP adresses are more like uranium than gold. --Michael Dillon
Re: [Fwd: RE: Kremen VS Arin Antitrust Lawsuit - Anyone have feedback?]
> > The reason that ARIN allocations are not property is > > that pre-ARIN allocations were not property. ARIN is > > merely continuing the former process with more structure > > and public oversight. Are telephone numbers property? > IP addresses appear to be property - - read http://news.findlaw.com/ > hdocs/docs/cyberlaw/kremencohen72503opn.pdf. Given that domain names > are property, IP addresses should be property, especially in > California where are constitution states "All things of value are > property" 1. I searched that PDF and it says nothing whatsoever about IP addresses, therefore your statement above is not true. 2. The court didn't just say that domain names are property like anything else, he said that some of the laws regarding property apply to domain names. But others do not. 3. Domain names are delegated to people who pay money to register a domain name for their exclusive use forever as long as they maintain their renewal payments. 4. IP addresses are assigned to organizations who have a JUSTIFIED technical requirement for those addresses in their network. Most addresses can only be used as long as they remain connected to the same upstream network. PI addresses can only be used as long as they are technically justified. If an organization sells their network or shuts it down, then they can no longer keep their IP addresses and there are hundreds of cases where those addresses have gone back to ARIN to be allocated to other networks. > Why is it that they involve lawyers, > ask you all your customers names and etc... This is more information than > I think they should be requiring. Any company that wishes to engage in > business as an ISP or provider in some capacity should be granted the > right to their own ip space Look at this page: http://www.arin.net/cgi-bin/member_list.pl Every one of those organizations has disclosed to ARIN all their customer names, etc... That is the way things are done. If you don't want to play ball like the rest of us, then you are not going to get IP addresses. That's the simple truth. We have a level playing field and you are asking for special privileges that other organizations don't feel are necessary. --Michael Dillon
RE: Kremen VS Arin Antitrust Lawsuit - Anyone have feedback?
>Even if you assume that allocations made by ARIN are not property, it's > hard to argue that pre-ARIN allocations are not. They're not subject to > revocation and their grant wasn't conditioned on compliance with policies. The reason that ARIN allocations are not property is that pre-ARIN allocations were not property. ARIN is merely continuing the former process with more structure and public oversight. Are telephone numbers property? In any case, since the conditions of the pre-ARIN allocations were all informal, unrecorded and largely verbal, nobody can prove that there was any kind of irrevocable grant. --Michael Dillon
Re: [Fwd: Kremen VS Arin Antitrust Lawsuit - Anyone have feedback?]
> Since the public policy meetings and mailing lists where > consensus is judged > are open to any interested party, it is very hard to view this as an > anti-competitive act in my > opinion. Kremen filed the suit on April 12, 2006. That is the last day of the ARIN public meeting in Montreal. I was at the Montreal meeting and Kremen never appeared publicly there to question ARIN's actions. It make me think that he did not make a reasonable attempt to resolve the situation out of court. --Michael Dillon
Re: [Fwd: Kremen VS Arin Antitrust Lawsuit - Anyone have feedback?]
> > Your statement about preferential treatment is factually > > incorrect. Larger ARIN members do not get larger allocations. > > It is the larger network infrastructures that get the larger > > allocations which is not directly tied to the size of the > > company. Yes, larger companies often have larger infrastructures. > > And that's the point: A company that is established gets preferential > treatment over one that is not; that is called a barrier to entry by the > anti-trust crowd. You need to understand the basics of networking to see that this is NOT preferential treatment but is instead even-handed treatment. You see, a network is a collection of devices interconnected with circuits. Each point where a circuit connects to a device is called an interface. Devices may have more than one interface. Typically, the devices used by network operators have many interfaces. IP addresses are numbers used to uniquely identify such interfaces and the Internet Protocol (IP) requires that these numbers be assigned in a structured manner. It is then obvious that larger networks have more interfaces and therefore can TECHNICALLY justify more addresses. This is even-handed treatment even though small companies end up with less addresses than large companies. >You may feel that such a barrier is justified and > fair, but those on the other side of it (or more importantly, their > lawyers) are likely to disagree. Yes, lawyers do not understand networks. No doubt some of them will read the above text and begin to get a glimmer of understanding. > Of course it's directly connected; all you have to do is look at the > current fee schedule and you'll see: > > /24 = $4.88/IP > /23 = $2.44/IP That is completely untrue. ARIN's web page here http://www.arin.net/billing/fee_schedule.html says nothing of the sort. In fact, ARIN's annual fees are structure so that organizations which have a larger transaction volume pay a larger fee. These transactions could be IP address applications or SWIP transactions or in-addr.arpa traffic. The size categories are just a rough rule of thumb for categorizing organizations that has been accepted by the ARIN members themselves. > So, just between the two ends of the fee schedule, we have a difference > of _two orders of magnitude_ in how much an registrant pays divided by > how much address space they get. Large organizations get their allocations bit by bit, applying for 3-6 months requirements at a time. Small organizations may have only a single allocation. > Besides the above, Kremen also points out that larger prefixes are more > likely to be routed, therefore refusing to grant larger prefixes (which > aren't justified, in ARIN's view) is another barrier to entry. Again, > since the folks deciding these policies are, by and large, folks who are > already major players in the market, it's easy to put an anticometitive > slant on that. Routability decisions are not made by ARIN. If anyone is unhappy with routability they should be suing those organizations which recommend route filtering. But they would have to prove that the route filtering is not technically justified which will be difficult when all the expert witnesses are on the other side. --Michael Dillon
RE: [Fwd: Kremen VS Arin Antitrust Lawsuit - Anyone have feedback?]
> 3) What's wrong with treating assignments like property and setting > up a market to buy and sell them? There's plenty of precedent for this: > > Mineral rights, mining claims, Oil and gas leases, radio spectrum. Before you start making inferences from an analogy, you had better be sure that you have the right analogy. IP addresses are not like any of the things that you mention. They are like phone numbers which also are not property and also managed by a central admin function NANPA. --Michael Dillon
Re: [Fwd: Kremen VS Arin Antitrust Lawsuit - Anyone have feedback?]
> The debate there will be around the > preferential treatment that larger ARIN members get (in terms of larger > allocations, lower per address fees, etc), which Kremen construes as > being anticompetitive via creating artificial barriers to entry. That > may end up being changed. Your statement about preferential treatment is factually incorrect. Larger ARIN members do not get larger allocations. It is the larger network infrastructures that get the larger allocations which is not directly tied to the size of the company. Yes, larger companies often have larger infrastructures. However, ARIN gives the addresses based primarily on the number of hosts attached to the network infrastructure. It has been argued in the past that ARIN's policies are prejudiced AGAINST larger organizations because the rules do not properly allow for the scaling overhead necessary due to the complexity of larger networks. As for fees, there are no per-address fees and there never have been. When we created ARIN, we paid special attention to this point because we did not want to create the erroneous impression that people were "buying" IP addresses. The fees are related to the amount of effort required to service an organization and that is not directly connected to the number of addresses. --Michael Dillon (no longer in any official ARIN capacity. Just another member)
Re: [Fwd: Kremen VS Arin Antitrust Lawsuit - Anyone have feedback?]
> I am looking for anyone who has input on possibly the largest case > regarding internet numbering ever. This lawsuit may change the way > IP's are governed and adminstered. Comments on or off list please. My personal opinion is that this is yet another example of ignorance leading to anger leading to a stupid waste of court time. The case is filled with incorrect statements of fact which ARIN can easily demolish. But at the bottom line, these people are complaining because ARIN didn't let them use some IP addresses that were assigned to a different company. Since IP addresses are basically available free from any ISP who sells Internet access services, this seems like a severe error in judgement on the part of the plaintiff. A smart businessperson would have used the free IP addresses to keep their business online even if they did decide to dispute ARIN's decision. But in the end, IP addresses are not property, therefore they cannot be assets and cannot be transferred. They can only be kept if they are in use on network assets which are transferred and which continue to be operational. And even then, most people have no choice as to which specific address block they use. They simply take what the ISP gives them. I personally suspect that ARIN will have this thrown out of court in fairly short order. Even if it did go much further, the parallels with NANPA would see it fade away quite quickly. This discussion really belongs on http://www.groklaw.net/ where I note it has not yet appeared. Perhaps another indication that this is a tempest in a teapot. --Michael Dillon
RE: Router / Protocol Problem
> Apparently some how this connection is being > >matched via NBAR for good old Code Red. > Best moved to cisco-nsp. What!? Network operator discovers that measures taken to mitigate an old network security measure, long past their sell-by date, are now causing random grief. Seems to me like bang on topic for NANOG. What other such temporary mitigating measures are still in place long after the danger has passed. Note, that Code RED was a both an application vulnerability and a network DDoS. Even though there are likely still many hosts running the vulnerable application, the number is not sufficient to cause another massive DDoD and measures taken to protect against this particular peculiar DDoS, really don't have a good technical reason to remain in place. This is probably also another instance of the well-known ops problem: We know how to get stuff deployed but we can't undeploy stuff because we are too busy deploying other stuff. --Michael Dillon
Re: Spain was offline
> > I can't get a TLD zone? But back to the root servers. Are you > > agreering with me that if I announce F and I root's netblocks > > inside of my own network that everyone would be ok with that? > Who is responsible if this set-up fails? > > Who is responsible if it lies? > > Who is likely to get blamed for any failures? > > Would this require explicit consent from all customers > subject to such treatment? > > Would this require a possibility for each custoemr to opt out > of such a scheme? Aren't all of these questions private issues between the private network operator and their customers? The same thing applies to companies who use IP addresses inside their private networks that are officially registered to someone else. This is a fairly common practice and yet it rarely causes problems on the public Internet. Since Internet network operators are generally not regulated in how they operate their IP networks, it seems to me that the people who say that it is not proper to announce root netblocks in a private network are really calling for network regulation by an external authority. > And - ah yes - what particular problem does such a set-up solve? It seemed to me to be a theoretical question not intended to solve a particular problem. However, theoretically, a network that sources a lot of DDoS traffic to root servers could do this to attract the traffic to their local copy of the root server in order to analyze it. Theoretically, this is something that would be enabled by the hypothetical situation described above. --Michael Dillon
Media reports (was: Spain was offline)
> [EMAIL PROTECTED]> host www.red.es > www.red.es is an alias for web.red.es. > web.red.es has address 194.69.254.50 > > No idea what happened, and I don't read spanish, According to red.es, they believe that a possible hardware failure caused a file to be corrupted during the update. When this was discovered, they shifted to a backup file. Details here for those who do read Spanish: http://www.noticias.info/asp/aspComunicados.asp?nid=214866&src=0 --Michael Dillon
Re: Spain was offline
> You seem to be suggesting that ISPs run stealth slaves for these > kinds of zones. Not really. In today's world such simplistic solutions don't work. > For zones that are being made available on anycast servers, ISPs may > be able to lobby/pay the zone operator to install an anycast instance > in their network. However, in general, the days of ISPs being able to > set these things up on their own and see benefit from them are past, > in my opinion. I believe that there are still some things that ISPs can do which cannot simply be bought on the market. For instance, most ISPs runs simple caching servers for their DNS queries where they keep any responses for a short time before deleting them. It's so simple that it is built into DNS relays as an option. An ISP could run a modified DNS relay that replicates all responses to a special cache server which does not time out the responses and which is only used to answer queries when specified domains are unreachable on the Internet. For instance, if you specified that all .es responses were to be replicated to the cache and that your DNS relay should divert queries to the cache when .es nameservers are *ALL* unreachable, then the impact of this type of outage is greatly reduced. You could specify important TLDs to be cached this way as well as important domains like google.com and yahoo.com. The actual data cached would only be data that *YOUR* customers are querying anyway. In fact, you could specify that any domain which receives greater than x number of queries per day should be cached in this way. The volume of data cached would be so small in todays terms that it only needs a low-end 1U (or single blade) server to handle this. Since nothing like this exists on the market, the only way for ISPs to do this is to roll their own. Of course, it is likely that eventually someone will productize this and then you simply buy the box and plug it in. But for now, this is the type of thing that an ISP has to set up on their own. --Michael Dillon
Re: Spain was offline
> Do you know how to contact your network provider without looking up > e.g. www.example.com (network provider web site)? IS TCPWRAPPER > configured to lookup names before allowing an operator login on a > critical server? Do you know your name servers IP addresses? If > the PSTN phone numbers don't work, do you have a INOC-DBA phone? If > the INOC-DBA phone numbers don't work, do you have a PSTN phone number? Do you have your own mirrors of TLDs that are important to your users, i.e. .com, your .xx country domain, etc.? --Michael Dillon
Re: Fanless x86 Server Recommendations
>...but the fanless chips are > not always as fanless as you might like. I've seen a number of them come > back well fried. Fanless doesn't just mean "no fans to break down". It also means "well-ventilated installation required". Maybe you can find a datacenter with so many hot bladeservers that they can't fill their racks. Then you could ask them to give you 8U cheap at the bottom of a rack and mount your servers vertically for maximal airflow. ;-) --Michael Dillon P.S. on the other hand, if there is enough demand for fanless server installations, maybe some datacenters will begin to offer vertically vented space at the bottom of their racks...
Re: Fanless x86 Server Recommendations
> > We're looking to acquire a couple small servers that can act as routers for > > us at remote locations. > >You may want to check out soekris. (www.soekris.com) This type of server is far more common nowadays than it was when Soekris launched their business. A Google search will lead you to dozens of fanless servers built around a VIA EPIA mini-itx board or one of AMD's GEODE chips. --Michael Dillon
Re: IP failover/migration question.
> It's actually a rather frustrating > situation for people who aren't big enough to justify a /19 and an > AS#, but require geographically dispersed locations answering on the > same IP(s). If the number of IPs you require is small, then you can probably solve the problem with IPv4 anycasting. Several people have built out distributed anycast networks but the problem is that they think IPv4 anycast is a "DNS thing". Therefore they don't sell anycast hosting services to people like you who need it. Of course, if you made them more aware of market demand, this could change. --Michael Dillon
Re: Who wants to be in charge of the Internet today?
> The Business Roundtable, composed of the CEOs of 160 large U.S. companies, > said neither the government nor the private sector has a coordinated plan > to respond to an attack, natural disaster or other disruption of the > Internet. While individual government agencies and companies have their > own emergency plans in place, little coordination exists between the > groups, according to the study. I don't believe that this is entirely true. I think that there is a lot of coordination between companies at an industry level, for instance the automotive industry or the financial services industry. This coordination doesn't get much visibility outside of the industry concerned but that doesn't mean that it isn't there. In fact, I strongly suspect that visibility of this coordination does not often reach the CEO level in these companies because much of the coordination is between specialist groups within the companies. Does your CEO know that you participate in NANOG? One might even venture to suggest that there is no point in coordinating emergency plans between companies who have little or no direct business relationships unless it is at a metropolitan level, i.e. New York area businesses, Los Angeles area businesses. After all, why should NY businesses plan for earthquakes and why should LA plan for a hurricane? --Michael Dillon
Re: DNSSEC in Plain English
> but it ain't the crypto. never has been. and it is not always > easy to explain math in plain english. so let's focus on where > work needs to be done. You and I are in violent agreement. The problem is in understanding whether or not the crypto under the hood really does provide a TRUSTABLE system. And that is more to do with policies and procedures. This is the stuff that I don't see explained in plain English so that the decision makers who rely on DNS can make a decision on DNSSEC. Ed Lewis pointed out two presentations which he claims have no crypto. However his own presentation at Apricot is laced with technical jargon including crypto. Stuff like "hierarchy of public keys", "DNSSEC data", "hash of the DNSKEY", "certificates", and so on. This is fine for a technical audience but it won't help explain the issue to the decision makers who spend the money. I understand how the crypto works to the extent that I believe it is technically possible for something like DNSSEC to work. However, I don't see an explanation of the policies and procedures that convinvces me that it DNSSEC really does work. The history of crypto-based security is filled with flawed implementations. --Michael Dillon --Michael Dillon
Re: wrt joao damas' DLV talk on wednesday
> actually, in a brilliant demonstration of fair use of copyrighted > lyrics, paul was quoting directly from the song about alice's > restaurant. well, actually, despite saying so, it's not much about > the restaurant at all. "and the restaurant is not called alice's > restaurant, that's just the name of the song." And this whole discussion about DNSSEC and DLV reminds me of a bunch of 8 x 10 glossy photographs with the circles and arrows and a paragraph on the back of each one. Just another case of American blind justice I suppose. Has anyone ever considered trying to come up with a way that these crypto projects could be explained in plain English? I think a lot of the problem with adoption of DNSSEC stems from the fact that most people who might make a decision to use it, haven't got a clue what it is, how it works, or whether it even works at all. And it's not their fault that they don't understand. It's the fault of a technical community that likes to cloak its discussions in TLAs and twisted jargon. --Michael Dillon
Re: Tracing network procedure for stolen computers
> Earlier this month my daughters Ibook was stolen, oh well that is life I > guess. > Anyway updated mail server software for full debug and IP log since noticed > that mail account was accessed yesterday. It's a UNIX machine. You own it. You know the password. If you had only set up an SSH server on it, you would now be able to log in and collect additional information about the current user. Interesting things can happen when intelligent devices find themselves stolen... http://www.evanwashere.com/StolenSidekick/ --Michael Dillon
Re: wrt joao damas' DLV talk on wednesday
> attending nanog wasn't an option. i hadn't realized that sitting in on > joao's talk so i could be there for Q&A equalled attendance, and so i > neither paid nor offered retroactively to pay. Sounds to me like your intent was to be a "Speaker" > do you really think i > should? (i asked everybody i met on site, and was universally told by > those i asked to "stop worrying about it".) If Merit had simply given you a "Speaker's Badge" then all this tempestuous teapot wouldn't have dribbled a single drop. --Michael Dillon
Re: IP failover/migration question.
> clear understanding as to what is involved in terms of moving the IPs, > and how fast it can potentially be done. I don't believe there is any way to get the IPs moved in any kind of reasonable time frame for an application that needs this level of failover support. If I were you I would focus my attention on maintaining two live connections, one to each data centre. If you can change the client software, they they could simply open two sockets, one for traffic and one for keepalives. If the traffic destination datacentre fails, your backend magic starts up the failover datacentre and the traffic then flows over the keepalive socket. And if you can't change the clients, you can do much the same by using two tunnels of some sort, MPLS LSPs, multicast dual-feed, GRE tunnels. The Chicago Mercantile Exchange has published a network guide that covers similar use cases. In the case of market data, they generally run both links with duplicate data and the client chooses whichever packets arrive first. Since market data applications can win or lose millions of dollars per hour, they are the most time-sensitive applications on the planet. http://www.cme.com/files/NetworkingGuide.pdf > When I desire to migrate hosts to the failover site, B would send a > BGP update advertizing that the redundant link should become > preferred, There is your biggest timing problem which is also effectively out of your control. By maintaining two live connections over two separate paths to two separate data centers, you have more control over when to switch and how quickly to switch. --Michael Dillon
Re: wrt joao damas' DLV talk on wednesday
> you were attending nanog without registering and paying? that is > rude. have you offered to pay retroactively? that would be the > honorable thing to do. > todd underwood +1 603 643 9300 x101 > renesys corporationchief of operations &security There was a similar comment from another Renesys employee on nanog-futures. Is it possible that this is some sort of commercial dispute between two companies, Renesys and ISC, who are not network operators, but who offer services to network operators? In any case, it doesn't seem to be on topic for the NANOG list. If Renesys really doesn't like ISC, why don't you sue him instead of whining on this list? --Michael Dillon
Re: Extreme Networks BD 6808 errors -- help to interpret.
> I have probably missed something, perhaps unwritten policy, and for that I > am sorry. I will not repeat my mistake. Please DO CONTINUE to discuss this on the list. Ignore all those messages of complaint. The only complaints that matter are those of the Mailing List Administrators whose names are listed here: http://www.nanog.org/listadmins.html The people who were complaining to you are not serious about network operations. They just want to keep it as a private club where only people who know the secret handshake can apply. However, in the 21st century, stable and reliable network operations are vital to the global economy. This means that we MUST openly discuss issues that arise in order to jointly solve the problems and to educate all parties involved, vendors, researchers and operators. It's OK to step on some toes and offend a few people. This is a rough and tumble business where you need to have a thick skin to survive. Perhaps the problem is that the COMPLAINANTS do not have a thick enough skin. --Michael Dillon
Re: 2006.06.06 NANOG-NOTES IDC power and cooling panel
> Dan Golding, 30 seconds, what would each person like > to see the folks on other side do to help. Did anybody mention cogeneration? At least some of that waste heat is turned into electricity which reduces the total consumption off the grid. http://www.polarpowerinc.com/products/generators/cogenset.htm And what about risks? As you increase the active cooling of your IDC, you reduce the ability to survive power outages. This is another reason to separate hot customers from cool. In the event of a power outage, your cool customers who have better heat engineering do not have to share fate with the lazy hot customers. Here I am referring to server customers who do have choices to use more efficient hardware, improve the power efficiency of their software, and do things like server virtualisation to reduce the CPU count in your IDC. The embedded systems industry has plenty of experience in reducing power consumption through both hardware and software improvements. --Michael Dillon
RE: Zebra/linux device production networking?
> I would be interested to know how many "software" (for want of a better > description) routers are in live production in this kind of environment > i.e. the 99.% Uptime variety, from speaking to people albeit > randomly in data centres it would seem to be more common than one might > expect. It is indeed very common. That is why there are several implementations of BGP and routing software available. These are used in dozens and dozens of commercial products some of which are sold as IP routers, plain and simple. In any case, 5 nines and 6 nines are not always what the marketing department claims. They often exclude planned maintenance periods so if you reboot once a week or you have a crash after changing a config, that doesn't count against the 5 nines. In addition, the 5 nines figure generally applies to the network, not to individual devices within it. Networks can be designed so that the failure of a device does not cause a network outage. This whole issue is so complex that you just can't make blanket recommendations. Even the biggest networks don't just buy and deploy big iron. They run every new router model and software release through an extensive battery of tests. Then they write operational guidelines telling people which features can be used in which situations. They do this to avoid crashes and network outages because the big iron (Cisco/Juniper) simply cannot provide that on its own. A smart small company can get excellent results from Linux routers (although I would take a serious look at FreeBSD or OpenBSD for this). Process is as important as hardware. --Michael Dillon
Re: Zebra/linux device production networking?
> First, a little background.. > My CTO made my stomach curdle today when he announced that he wanted to > do away with all our cisco [routers] and instead use Linux/zebra boxen. > We are a small company, so naturally penny pinching is the primary > motivation. It is primarily small companies that use zebra or Quagga or openbgpd or Xorp or the Click Modular Router project. There is more than one choice so do your research. The main drawback of all of these is that you cannot get PCI-bus cards that support some common circuit types and the PCI bus cannot handle switching high traffic volumes. Many people build and sell routers based on a PC server running UNIX. They work fine if they are no stretched beyond the role intended. Cisco routers are the same. Look at the limitations of the 2500/2600 series for instance. Some URLs of interest: http://www.read.cs.ucla.edu/click/ http://www.xorp.org/ http://www.openbgpd.org/ http://www.quagga.net/ http://www.zebra.org/ > Has there been any discussion (or musings) of moving towards such a > solution? I've seen a lot of articles talking about it, but I've not > actually seen many network operators chiming in. This tends to be a list focused on the cult of the BIG IRON, namely Cisco and Juniper. The people who use PC-based routers have their own hangouts. My main piece of advice is to seek out those hangouts and ask your questions there. > Here's the article that started it all (this was featured on /., so > likely you've read it already). Sorry, haven't seen these. --Michael Dillon
Re: Phantom packet loss is being shown when using pathping in connection with asynchronous routing - although there is no real loss.
> The only part that I don't get is that you can mtr to him without > packetloss. Although the path in-between may be different, the final hop > packetloss should exactly equal what he sees when mtring you. A round-trip > is a round-trip, and results should be identical regardless of who > originates. I can't think of any way this would be different unless echo > and echo-reply were being rate limited independently. If the time was different then the packet loss would be different. Perhaps the customer runs the tests during his busy period when he is concerned about making sure there is no delay. Then, later in the day, after his busy period is over he takes the time to contact his ISP. The ISP then runs some tests which show there is no packet loss at all. To be sure this is not happening, synchronize the tests and run simultaneously. Try tcptraceroute because this more accurately reflects the traffic that is flowing. http://michael.toren.net/code/tcptraceroute/ http://tracetcp.sourceforge.net/ is a windows tool that is similar. The open source tool LFT can be built to run on Windows under cygwin http://pwhois.org/lft/ but they have this warning on their page: Many people have complained about various problems on the Windows platform. Both LFT and the WhoB client compile and run well under Cygwin environments on Windows. Unfortunately, Microsoft's changes to the Windows IP stack (as of XP Service Pack 2) reduced their raw socket functionality significantly as part of their security bolstering process. These changes have effectively stopped LFT from working properly while using TCP. LFT's UDP tracing and other advanced features still work properly. For more information on Windows raw sockets, consult www.microsoft.com/technet/prodtechnol/winxppro/maintain/sp2netwk.mspx#EIAA This may have nothing to do with your MTR issue but it does make one wonder whether a Windows machine is safe to do performance testing. In any case, the LFT people think that their non-TCP features still work properly on Windows and this is a tool that you can also run on your end. Worth a try? --Michael Dillon
Re: Layer3 down?
> From my IT department: > > "It seems that the internet is having issues at a Layer3 Communications > router, this normally would not be a problem but L3 runs some routing for > the internet backbone. The techs at L3 are working on the problem and we do > not have an eta as to when it will be back up, this situation is completely > out of our control. You may experience short network outages or severe > latency." Thanks for posting my morning smile. :-) --Michael Dillon
Re: Are botnets relevant to NANOG?
> The motive is unclear because attacking, > for example, root servers, is an effort without some obvious economic > incentive Since when is advertising NOT a sign of an obvious economic incentive? > The DA report went through a large thread(s) to post statistics here > and I'm not sure why yours will be any better, or, just another set > of statistics which further de-sensitizes everyone to the problem. Stats by themselves can be boring. But making data available regularly and publicly will inevitably lead to some people doing analyses of this data and presenting those analyses to NANOG meetings. This will lead to wider understanding of what is going on and will provide raw material for getting management support for actions to solve the problem. --Michael Dillon
Re: Are botnets relevant to NANOG?
> for this community would trend analysis with the best of who is getting > better and the worst of who is getting worse and some baseline counts be > enough for this group to understand if the problem is getting better. Your 5-day numbers were very reminiscent of the weekly CIDR report. I think that if you clean it up for weekly submission then that would be useful to some. For instance, you only published data for two categories of ASN. Where is the tier-1 data? And numbers should cover a 7-day period, not 5 days. In addition, for each category you should provide a fixed cutoff. The CIDR report shows the top 30 ASNs. I think that the ideal would be a table including ASN category as one column and showing the top 50 ASNs. In addition, you should attempt to separate the dynamic addresses by some means or other and either add that as a separate column or else do a separate table. Since this would be posted weekly over a long period of time, it is best to put some thought into how to structure it so it remains relevant. Also, provide a URL where researchers can download more complete datasets, not just top 50. > I am suggesting that NANOG is an appropriate forum to publish general > stats on who the problem is getting better/worse for and possibly why > things got better/worse. I think few people will complain about a weekly posting of this nature. --Michael Dillon
Re: AS12874 - FASTWEB
> >> http://plany.fasthosting.it/dbmap.asp?table=Mappatura > > I take it that this means we can use any ip range allocated to Fastweb > > as if it were RFC1918 space, including the necessary border filters? > I'd personally contract to build a moat around their NOC for Homeland > Security reasons using as many backhoes as I could get on short notice. I would strongly advise against such actions. European governments take a dim view of terrorist activities and some countries such as Italy are particularly sensitive about this. I'm surprised that an American on an Internet operations mailing list would be promoting terrorist activity in another NATO member country. In any case, you can't CONTRACT to do this. The law does not consider an agreement to perform illegal acts to be a contract. The action you describe is clearly illegal, therefore it cannot be contracted for. --Michael Dillon P.S. this is NANOG, not IRC
Are botnets relevant to NANOG?
In recent discussions about botnets, some people maintained that botnets (and viruses and worms) are really not a relevant topic for NANOG discussion and are not something that we should be worried about. I think that the CSI and FBI would disagree with that. In a press release announcing the last CSI/FBI survey http://www.gocsi.com/press/20050714.jhtml the following statement appears: Highlights of the 2005 Computer Crime and Security Survey include: - The total dollar amount of financial losses resulting from security breaches is decreasing, with an average loss of $204,000 per respondent-down 61 percent from last year's average loss of $526,000. - Virus attacks continue as the source of the greatest financial losses, accounting for 32 percent of the overall losses reported. - Unauthorized access showed a dramatic increase and replaced denial of service as the second most significant contributor to computer crime losses, accounting for 24 percent of overall reported losses, and showing a significant increase in average dollar loss. So where do botnets come in? First of all, botnets are used to distribute viruses, the largest source of financial losses. Second, botnets are built on what the CSI calls "unauthorised access", the second largest source of loss. And denial of service, which used to be the 2nd largest, is also something that botnets do. Now NANOG members cannot change OS security, they can't change corporate security practices, but they can have an impact on botnets because this is where the nefarious activity meets the network. Therefore, I conclude that discussions of botnets do belong on the NANOG list as long as the NANOG list is not used as a primary venue for discussing them. One thing that surveys, such as the CSI/FBI Security Survey, cannot do well is to measure the impact of botnet researchers and the people who attempt to shut down botnets. It's similar to the fight against terrorism. I know that there have been 2 terrorist attacks on London since 9/11 but I don't know HOW MANY ATTACKS HAVE BEEN THWARTED. At least two have been publicised but there could be dozens more. Cleaning up botnets is rather like fighting terrorism. At the end, you have nothing to show for it. No news coverage, no big heaps of praise. Most people aren't sure there was ever a problem to begin with. That doesn't mean that the work should stop or that network providers should withold their support for cleaning up the botnet problem. ------- Michael Dillon Capacity Management, 66 Prescot St., London, E1 8HG, UK Mobile: +44 7900 823 672Internet: [EMAIL PROTECTED] Phone: +44 20 7650 9493Fax: +44 20 7650 9030 http://www.btradianz.com One Community One Connection One Focus
Re: ISP compliance < LEAs - tech and logistics
> > The guy wants to say, please raise your eyes above the horizon of your > > plate and view a not yet existing country named europe. Here our > > infrastructure is a lot more advanced and we have standardized a > > common eavesdropping api. > > We have? News to me. You missed a line later in his message: >Of course >nobody except the European Central Bank is allowed listening, but - >who cares? Sounds like typical lunatic ravings to me. I guess anything goes on this list now... --Michael Dillon
Re: private ip addresses from ISP
> > Does NANOG have a role in developing some best > > practices text that could be easily imcorporated > > into peering agreements and service contracts? > ... > > RFC 2267 -> RFC 2827 == Best Current Practice (BCP) 38 > RFC 3013 == BCP 46 > RFC 3704 == BCP 84 > Are these followed? No, the IETF BCP's are not followed and part of the reason is that they are not written by network operators but often by vendors and academics. The fact is that the collective of network operations people (in North America at least) does not have any agreed set of BEST OPERATIONAL PRACTICES. There are various camps that promote various sets of rules which are often overly simplistic and cannot be 100% adhered to in practice. What we really need is a forum to discuss best operational practices and resolve all these various differences in opinion systematically. The end result should be a set of best practices that people really can follow with no exceptions. Of course this means that the best practices must incorporate various exceptions to the simple rules and explain when those exceptions are and are not appropriate. So again, I ask the question: Is NANOG an appropriate forum to develop some best practices text that could be incorporated into service agreements and peering agreements by reference in the same way that a software licence incorporates the GPL by referring to it? --Michael Dillon
Re: ISP compliance < LEAs - tech and logistics [was: snfc21 sniffer docs]
> >The NANOG meeting archives are full of presentations as the result > >of very sophisticated network monitoring. Like most technology, > >it can be used for good and evil. You can't tell the motivation > >just from the technology. OK, so he says in a roundabout way that you are already paying for some sophisticated network monitoring and it probably won't cost you much to just give some data to the "authorities". > Sean, please drop this subject. You have no experience here and it's > annoying that you keep making authoritative claims like you have some > operational experience in this area. If you do, please do elaborate > and correct me. From what I understand from the folks at SBC, you > did not run harassing call, annoyance call, and LAES services. I would > appreciate a correction. Huh!?!?!? Are you saying that people should buzz off from the NANOG list if they change jobs and their latest position isn't operational enough? Are you saying that people should not be on the NANOG list unless they have TELEPHONY operational experience? What is the world coming to!? --Michael Dillon
Re: private ip addresses from ISP
> Proper "good net neighbor" egress filtering of RFC1918 source addresses > takes a number of separate rules. Several 'allows', followed by a default > 'deny'. Really? Do you have those rules on your network? Any reason why you didn't post the operational details on this operational list? Have you ever read your peering agreements or service contracts to see if filtering of RFC 1918 sourced traffic is specifically covered by them? If it is not covered by the contract, then why should your peers/upstreams filter it? Another good question is whether or not every service contract and peering agreement should contain unique text or whether there should be some community-developed best practices statement that could be plugged in by reference. For instance, software publishers can publish their software under the terms of the GPL without including the full text of the GPL verbatim in their software license. Does NANOG have a role in developing some best practices text that could be easily imcorporated into peering agreements and service contracts? --Michael Dillon
Re: Geo location to IP mapping
> I just tried that, says I'm 100 miles south of where I really am. That's > quite a long way out in a small country like England. I live in London and use BT Broadband. But geolocation shows me being in Ipswich up in East Anglia, a long way from London. I assume this is because the geolocation only knows that I use an IP address from a DHCP pool managed in Ipswich. They don't know anything about BT's own extensive network, and in the case of DSL using tunnels and DHCP servers, the real topology becomes entirely invisible. The end result is that most of England's population lives in Ipswich. Eat your heart out Alan Partridge. A few years ago, while working at a different company in London, I had a New York IP address because our company's internal network Internet gateway was in New York. Then they changed things around so that we used a gateway in London for all the European offices. But that meant that colleagues in France, Germany, etc... would all show up as being located in London. Nowadays, I use a VPN to work from home. The VPN software knows of multiple tunnel endpoint servers so if there is a problem with the UK server it fails over to a server in the USA. My IP address on the Internet comes from the NAT server at the Internet gateway. Depending on where the tunnel endpoint is, it could be a US address or a UK address. 100% accurate geolocation is not achievable but if you understand the issues then you can better make a decision how to apply geolocation services to your own problem. It may work well enough for some things. --Michael Dillon
Re: Geo location to IP mapping
> I can solve the visualization part and the GIS issues. But comes down > to the accuracy of the geo-ip database in the end. According to the Brand X localisation database which was rated tops in the Brand Y Web Magazine survey in 2005, our top customers are located in these cities. Who said marketing is not for techies? --Michael Dillon
Re: Geo location to IP mapping
> As a major caveat, all geolocation services do have some degree of > inaccuracy, because the sources of data are very diverse. (Some ISPs > provide complete subnet maps to MaxMind and other providers, whereas > some data is scraped from WHOIS or provided by inference from > end-users.) And some organizations run their own internal networks across international borders. In other words, knowing that subnet X is allocated to company Y who has a 300 meg Internet connection in city Z, does not mean that all the users of that connection are also in city Z. They could be scattered around the world. This is why some companies use other sources of data to infer the location, i.e. if users of an IP address prefer yahoo.fr to yahoo.com, then that is one datapoint in favour of them being located in France. If you understand the principles of RBL weighting then you will get the idea. --Michael Dillon
Re: MEDIA: ICANN rejects .xxx domain
> > But there's no technical advantage of a hierarchical system over a > > simple hashing scheme, they're basically isomorphic other than a hash > > system can more easily be tuned to a particular distribution goal. > > Amazing how many experienced people seem to be saying this isn't possible, > given there are already schemes out there using flat namespaces for large > problems (e.g. Skype, freenet, various file sharing systems). Most of these > are also far more dynamic than the DNS in nature, and most have no management > overhead with them, you run the software and the namespace "just works". According to your description, this is a hierarchical naming system. At the top level you have Skype, freenet, etc. defining separate namespaces. Because DNS was intended to be a universal naming system, it had to incorporate the hierarchy into the system. > However I think the pain in DNS for most people is the hierarchy, but the > diverse registration systems. i.e. It isn't that it is delegated, it is that > delegates all "do their own thing". Seems to me that this is part of the definition of "delegate". Some would say that this makes for a more robust system than a monolithic hierarchy where everyone has to toe the party line. > I've always pondered doing a flat, simple part of the DNS, or even > an overlay, > but of course it needs a business model of sorts. It has been tried at least twice and failed. http://www.theregister.co.uk/2002/05/13/realnames_goes_titsup_com/ http://www.idcommons.net --Michael Dillon