September 1996 - ripe-list - mailman.ripe.net

Re: Top-50 Report
by Tony Barber 30 Sep '96

30 Sep '96

Tony Bates wrote: > > >This is a list of the "Top 50" players who if CIDRizing at the AS level >could make a significant gain in the reduction of the size of Internet >routing tables. This may be an over-estimation but it is hoped that >this can act as an incentive for the "Top 50" and others to look at >their CIDR capability. This is a revival of report that used to be >posted in the early days of CIDR deployment (we even had a cidrd list >and working group then). > >This looks purely at the classful routes in the system and shows what gain if >cidrizing at the AS-level could be made by forming an aggregate. > > --Tony > >P.S. Same caveat about AS name mappings not working quite right yet. >P.P.S. Unless there are loud objections I plan to automate this again soon. > The above is an excellent idea. Perhaps if you could mail the output to nanog and ripe-list the peer pressure would help and in some cases inform LIRS/ISPs what they may unknowingly be doing wrong. RIPE did this earlier inthe year for the European registries and it worked. Is there an APNIC equivelant ? It would also be a good idea IMHO if the Global registries (APNIC RIPE INTERNIC)had email lists, alongthe lines of the ones RIPE runs, which it encouraged new 'customers' to join. This may be the nearest we get to a global isp email list. Regards Tony

3 2

unsubscribe
by root＠inco.com.lb 27 Sep '96

27 Sep '96

unsubscribe

1 0

unsubscribe
by Jean-Philippe_Chenu＠grenoble.hp.com 27 Sep '96

27 Sep '96

unsubscribe

1 0

Re: Top-50 Report
by Paul Ferguson 26 Sep '96

26 Sep '96

At 11:48 AM 9/26/96 +0100, Tony Barber wrote: >The above is an excellent idea. Perhaps if you could mail the output to >nanog and ripe-list the peer pressure would help and in some cases inform >LIRS/ISPs what they may unknowingly be doing wrong. >RIPE did this earlier inthe year for the European registries and it worked. > >Is there an APNIC equivelant ? > Yes -- AP Ops <apops(a)apnic.net> - paul

1 0

The Limerick WG Strikes Back
by Schweikhardt＠RUS.Uni-Stuttgart.DE 26 Sep '96

26 Sep '96

Hello *, many of you will already have seen the recent ACTIVITY REPORT of the Limerick WG. Thank you very much, Roderick, for compiling all that information. While idling on the train home, I used the spare time to compose a few more limericks which I think might be good for a laugh (or two). limerick: LIM-SYNFLOOD descr: On syn flooding attacks text: A Bad Guy's (TM) host in Kent text: Bazillions of SYN packets sent text: Until buffers starved text: And syslogds barfed. text: We hope this guy soon will repent. author: JS268-RIPE source: RIPE limerick: LIM-WG-CANCELLED descr: Limerick WG report cancellation text: The Limerick group worked all night text: To get all those limericks right. text: But their presentation text: Was dropped -- what frustration! text: Hey folks, wasn't THAT impolite? author: JS268-RIPE source: RIPE All standard smileys apply. No offense meant. [I'm wondering why I'm in the networking business. In my next life I'll probably be a poet.] Jens Schweikhardt -- SIGSIG -- signature too long (core dumped)

1 0

subscribe
by Wallace Ng 26 Sep '96

26 Sep '96

subscribe

1 0

Limerick WG report
by Roderik Muit 25 Sep '96

25 Sep '96

Unfortunately, the RIPE chair decided that there was too little time left for a presentation of the Limerick Working Group. So this time it will have to be done on the RIPE mailinglist. ACTIVITY REPORT After the deployment of the limerick object in the RIPE Database, which was in effect the creation of the first LR (Limerick Registry) at RIPE, there are no proposals for new activities to be started up in the Limerick area. There are however some issues that have come up regarding the functioning of the LR: - The LR is currently still not in good shape because old-style limerick submissions (at the moment stored at ftp://ftp.ripe.net/ripe/limericks) are not incorporated in the LR. Converting the old style submissions to limerick templates in the RIPE Database is outside the direct scope of the RIPE NCC activity plan, and also there have been no volunteer efforts or separate funding for this, so incorporating the old data in the LR is done on a very low priority basis. New submissions will however be stored in the LR. - I have not heard any suggestion so far from the secret Limerick Working Group on a uniform naming scheme of the limericks. (This might be the effect of nobody knowing who are actually the members of this working group.) Currently, limerick submissions that are approved by the Limerick Working Group are stored in the LR with the following name: LIM-xx-number, where xx is either the number of the RIPE meeting or the nic-handle of the author. This is however not a standard, so it does not provide for a real good way to search for limericks in the LR. (I would like to point you to the following options that are the best alternative so far when trying to find a certain limerick: - inverse lookup; try 'whois -h whois.ripe.net -i author <nic-hdl>' for limericks from a certain author, or 'whois -h whois.ripe.net -i admin-c LIM1-RIPE' for all limericks approved by the Limerick Working Group. - a copy of the full LR is stored at: ftp://ftp.ripe.net/ripe/dbase/split/ripe.db.li.gz) - All limericks approved by the Limerick Working Group are ofcourse protected by a maintainer object to protect people from tampering with them. However, because of the lack of any hierarchy in the limerick space, hierarchical authorisation cannot currently be deployed. There were vague suggestions of creating a 'root limerick object' which would provide for hierarchical authorisation, but no definitive proposals from the Limerick Working Group have been seen yet. This has the effect that anybody can create limerick objects without anybody being notified. A (manual) investigation of the current state of the LR shows that this is currently not a problem, although at a later stage this manual checking does not scale well. Only one limerick has been found that seems to have been inserted without the Limerick Working Group knowing: limerick: LIM-NEWOBJECT descr: Wait until Wilfried hears about this! text: An object appears from the blue text: Defined by goodness knows who text: There's no doubt a feeling text: Of insider dealing text: If the database working group knew. admin-c: MN131 author: MN131 remarks: No comment notify: mnorris(a)hea.ie changed: mnorris(a)hea.ie 960513 source: RIPE NEW SUBMISSIONS A clear pattern can be seen in the number of submissions of limericks. It looks somewhat like this: Number of limericks 5 | # 4 | # # # # 3 | # # # # 2 | # # # # 1 | # # # # # # 0 |######################################## +---------------------------------------- Dec Jan Feb Mar Apr May Jun Jul Aug Sep -> time 1995 1996 Most RIPE members will see a clear parallel with the RIPE meetings. The irregularities in this pattern were all caused by a sudden raise of activity emerging inside British Telecom. As usual, the number of submissions, via various channels, reached its peak at a tuesday night. It seems we are getting more activity, and quality is also quite high, which breaks with the trend in 1995 where one person took credit for almost all submissions. New submissions follow below. (JS268-RIPE == Jens Schweikhardt, by the way.) limerick: LIM-NT13-1 descr: RIPE24 dinner text: At a RIPE meeting, held in Berlin, text: The meal was unusual, the din text: Of the two serenaders text: Whose repertoire made us text: All smile and want to join in admin-c: LIM1-RIPE author: NT13 mnt-by: LIM-MNT changed: limerick-wg(a)ripe.net 960529 source: RIPE limerick: LIM-NT13-2 text: A document RIPE-181 text: Tells how everything ought to be done text: There are objects galore text: Heaven knows what they're for text: 'Cos they're all out of date, every one. admin-c: LIM1-RIPE author: NT13 mnt-by: LIM-MNT changed: limerick-wg(a)ripe.net 960529 source: RIPE limerick: LIM-NT13-3 text: The document RIPE-104 text: Goes on longer than ever before text: If you want an address text: And your network's a mess text: Then Daniel will get very sore admin-c: LIM1-RIPE author: NT13 remarks: Great limerick, but the draft document RIPE-104++ remarks: is now called RIPE-140... mnt-by: LIM-MNT changed: limerick-wg(a)ripe.net 960529 source: RIPE limerick: LIM-NT13-4 text: BGP is a protocol smart text: On Ciscos its really an art text: If you screw up your MED text: And your router goes dead text: Then you're better off with a restart admin-c: LIM1-RIPE author: NT13 mnt-by: LIM-MNT changed: limerick-wg(a)ripe.net 960529 source: RIPE limerick: LIM-NT13-5 descr: On the subject of joining the limericks-wg text: The limericks writers at RIPE text: Are nervous and shun any hype text: You can't get on their list text: And find out what you've missed text: 'Cos none of them knows what to type admin-c: LIM1-RIPE author: NT13 remarks: The limericks-wg members have no comment. Their identity remarks: remains a big mistery... mnt-by: LIM-MNT changed: limerick-wg(a)ripe.net 960604 source: RIPE limerick: LIM-NT13-6 text: If your secret just musn't get out text: And it's email you're worried about text: There's a method you see text: You can use PGP text: And send it without any doubt admin-c: LIM1-RIPE author: NT13 mnt-by: LIM-MNT changed: limerick-wg(a)ripe.net 960716 source: RIPE limerick: LIM-RIPE25-1 text: If we did not write a limerick text: Mike Norris might feel a little sick text: He left without a trace text: No one can take his place text: Nontheless we tried a rhyme quick admin-c: LIM1-RIPE author: LIM1-RIPE remarks: Author is actually a group: remarks: Brian Renaud, Ambrose Magee, Joachim Schmitz, Juergen Rauschenbach remarks: Produced at RIPE dinner ofcourse mnt-by: LIM-MNT changed: limerick-wg(a)ripe.net 960925 source: RIPE limerick: LIM-RIPE25-2 descr: Referring to an announcement on ripe-list shortly before RIPE25 text: Peter has a SONET line text: He thinks this is mighty fine text: He says the cell text: must go to hell text: I wish his bathroom was mine admin-c: LIM1-RIPE author: LIM1-RIPE remarks: Author is actually a group: remarks: Brian Renaud, Ambrose Magee, Joachim Schmitz, Juergen Rauschenbach remarks: Produced at RIPE dinner ofcourse mnt-by: LIM-MNT changed: limerick-wg(a)ripe.net 960925 source: RIPE limerick: LIM-RIPE25-3 descr: Referring to the Database Workshop at RIPE25 text: Our Ignorance was rendered moot text: by slides drawn amazingly cute text: To demonstrate series text: of database queries text: Ms. Orange used all sorts of fruit admin-c: LIM1-RIPE author: JS268-RIPE mnt-by: LIM-MNT changed: limerick-wg(a)ripe.net 960925 source: RIPE limerick: LIM-RIPE25-4 text: There once was a group with potential text: It's growth has been deemed exponential text: The database working text: and no more bugs lurking text: It's problems are purely financial admin-c: LIM1-RIPE author: JS268-RIPE mnt-by: LIM-MNT changed: limerick-wg(a)ripe.net 960925 source: RIPE

1 0

Re: Very fast IP, no ATM...
by Guy T Almes 24 Sep '96

24 Sep '96

Juha, Let me not speculate about all the various scenarios where the strengths and weaknesses of IP over ATM vs IP over SONET might vary. I addressed only the trans-oceanic very high delay-bandwidth product links, where I think the case is particularly strong. -- Guy At 11:11 AM 9/24/96 +0300, Juha Heinanen wrote: >when we talk about a link between two routers that will be saturated >immediately, atm doesn't give advantage over sonet. the situation >becomes different, when one wants to build a mesh network where the >traffic load betweeen the various nodes of the mesh varies and the >physical interface speed is fast enough to handle traffic between >several nodes simultaneously. > >-- juha >

1 0

Re: Very fast IP, no ATM...
by Paul Christ 24 Sep '96

24 Sep '96

On Sep 23, 5:41pm, Christian Huitema wrote: > Subject: Re: Very fast IP, no ATM... > > Did anybody in the `integrated services packet network' camp > > claim: ************************************************************************ * * * We can run the world-wide telephone system say by IPv6 * * > > over SDH? * * > * * > What make you believe we could not ? * * > * * > -- * * > Christian Huitema * * * ************************************************************************ Hereby I formally declare Christian THE Ultimate Internet Prophet Don't miss to deposit his visionary statement in a safe place. Artigst Paul Christ :-)

2 1

Re: Very fast IP, no ATM...
by Guy T Almes 24 Sep '96

24 Sep '96

Paul, I welcome your observation. The role of ATM in the Internet is a properly controversial one, and I'm sure the final page is not written. My point was intended to be narrow. I assume that the purpose of the trans-oceanic links is to effectively carry Internet traffic of a nature similar to usual Internet traffic. Given the very high price generally attached by carriers to T3 (and higher speed) leased lines, one is tempted to use ATM (particularly if it's priced lower). Many of the problems in the use of ATM to support Internet traffic become *especially* bad when one tries to support large delay-bandwidth products with traffic that is often quite bursty. This is often the case with such projects at the link that Peter described. When you consider a number of successes in the use of ATM in the Internet, such as the use of ATM within the MCI backbone here in the states, you will see that: - IP over ATM *can* be done correctly, particularly if one avoids overcommitting the bandwidth of the link, and - one often finds that IP over ATM is used when it's the only way to provision links faster than T3. One of the things that pleases me about the OC3 link that Peter announced this morning is that we may now find that there are (at least) two ways that engineers can consider in provisioning links faster than T3: IP over ATM and IP over SONET/SDH. As both alternatives become mature, we can begin to really compare which works best in various cases. Regards, -- Guy p.s.: I was specifically not trying to address the issue of 'world- wide telephone systems'. At 08:50 PM 9/23/96 +0200, Paul Christ wrote: >Hi, all of you, > >ok, congratulation, 'no ATM' - but: > >Don't you feel sometimes in all this fine argumentation against ATM >a certain intellektuelle Unredlichkeit (intellectual dishonesty)? > >Why? > >'ATM said' from the beginning: We can do everything - the >telephone system, eventually the video distribution and data. > >Did anybody in the `integrated services packet network' camp >claim: We can run the world-wide telephone system say by IPv6 >over SDH? > >(or would this be a non-goal?, or do we say, ATM also > wouldn't run the phone system, and so on; > years ago we had IBM's > packet transfer mode ... and so on) > > >Yours > > >Paul Christ > > >------------------------------------ >On Sep 23, 10:26am, Guy T Almes wrote: >> Subject: Re: Very fast IP, no ATM... >> Peter, >> Congratulations. This pleases me greatly. >> The success of the January 1994 trans-pacific T3 link from California >> to Hawaii by ANS was, in part, also due to taking advantage of direct >> IP over T3 frames. >> -- Guy >> p.s.: have you any measures of IP-level throughput over this link? >> >> >> At 03:43 AM 9/23/96 DST, Peter Lothberg wrote: >> > >> >On Monday 23 September 1996 00:15Z.. >> > >> >Was the worlds first transatlantic 155Mbit native IP service brought into >> >operation by Sprint/USA and Tele2/Sweden. >> > >> >The circiut wich is part of Sprintlink and ICMnet runs between the >> >NY-Nap in Pennsauken, NJ, USA and Tele2 in Sweden and uses cisco >> >packet-over-sonet/sdh technology. (Native IP over SDH/VC4) >> > >> >In 1995 the same team from Sprint and Tele2 brought up the worlds first >> >transatlantic E3 service between the same endpoints. >> > >> >For more information; >> > >> >Sprint: Tricia Schibler, +1 703 904 2042, ><tricia.schibler(a)qm.sprintcorp.com> >> >Tele2: Olle Wallner, +46 8 5626 4058, <wallner(a)swip.net> >> > >> >--Peter >> > >> >>-- End of excerpt from Guy T Almes > >

2 1