- db-wg - mailman.ripe.net

Announcement - Database Consistency and Statistics Project
by Can Bican 14 Jan '02

14 Jan '02

Dear Colleagues, [Apologies for multiple messages]. The new version of database consistency project, along with whois server usage statistics have been released on the RIPE NCC's web pages, with the following url: http://www.ripe.net/db/dbconstat/ The goal of this project is to find the objects with consistencies not enforced by the database, and provide reports and statistics to help users correct them. It also aims at checking for the necessary protection of objects and tracing the query and update statistics of the RIPE whois server to analyze the current trends of using the database. The goals of the consistency project do not include, for example, reality checks for inetnum objects, but it addresses only internal inconsistencies in the database, such as invalid references. Here are the types of inconsistencies which are addressed in this project: * Inetnum object related: * Inetnum objects without status attribute, * Assigned inetnum objects without allocations, * Overlapping inetnum objects, * Inetnum objects with suspicious boundaries to form a network, * Assigned inetnum objects from assignments, * Inetnum objects with wrong PA or PI status, * Wrongly allocated inetnum objects. * Person object related: * Person objects with multiple identical copies, * Person objects which do not have any references to them, * Person objects not having a valid nic-handle, * Objects which have references to invalid nic-handles. * Maintainer object related: * Maintainer objects which do not have any references to them. * Protection related: * Weak authentications schemes for maintainers. * Inetnum objects with no mnt-lower attribute. Reports about individual inconsistencies are sent only to the maintainers of objects. If you'd like to receive the report for your inconsistencies, send an e-mail to auto-dbcon(a)ripe.net containing the maintainer name in its body, and the appropriate authentication. * If the authentication of your maintainer is MAIL-FROM, mail should be sent from the appropriate e-mail address, and should contain the following in its body: mntner: <maintainer_name> * If the authentication of your maintainer is CRYPT-PW, mail should contain the following lines in its body: mntner: <maintainer_name> password: <password> * If the authentication of your maintainer is PGPKEY, mail should contain the following line and should be signed with the appropriate key: mntner: <maintainer_name> If you supply an existing maintainer and the proper authentication, the report of the inconsistencies related to the maintainer will be mailed back to you. For questions, comments and suggestions, please contact ripe-dbm(a)ripe.net. -- Can Bican RIPE NCC

1 0

Some RIPE Database updates were processed twice
by Andrei Robachevsky 14 Jan '02

14 Jan '02

Dear Colleagues, Due to misconfiguration of our main whois server some 300 updates were not processed correctly. We needed to rollback the database and process them again. Those affected may receive acknowledgement and notifications from the RIPE Database twice. In such case please ignore the first message. The service now is back to normal. We are very sorry for the inconvenience. Regards, Andrei Robachevsky DB Group Manager RIPE NCC

3 3

draft agenda (v1), for DB-WG meeting, RIPE 41, Amsterdam
by Wilfried Woeber, UniVie/ACOnet 13 Jan '02

13 Jan '02

Dear WG list members, attached please find the 1st (initial) draft: Proposed Agenda for the DB-WG Meeting at RIPE #41 in Amsterdam, NL. Please expect some minor changes/adjustments/re-sequencing closer to the meeting date. Looking forward to see many of you in Amsterdam a few days from now, Wilfried. ______________________________________________________________________ Proposed Agenda for the Database Working Group Meeting RIPE-41, Jan. 14-18 2002, Amsterdam, NL 1. Draft Thursday, Jan 17th, 2002 09:00-12:30 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ A. Administrative stuff (WW, chair) . volunteering a scribe . circulate list of participants . agenda bashing . status of minutes B. Actions Items . status of actions C. DB update (RIPE NCC, Andrei R.) 10 min . operations report D. DB software developments (RIPE NCC, N.N.) 30 min . bug fixes, re-writes, platform compatibility . meta-data proposal E. DB consistency and Statistics project (RIPE NCC, N.N.) 15 min F. IRT object implementation (RIPE NCC, N.N.) 15 min . implementation report, last-minute adjustments . how to use and promote the new features? G. Universal Whois Project (Andrew Newton, VeriSign App. Res.) 30 min . Needs of addressing information as it relates to domains in whois. . Overview of two technical solutions. . Unification of the addressing and domain databases into a more coherent "strategy". H. Modification of changed: attribute (tenative) I. "organisation:" Object re-visited (tentative) Y. Input from other WGs . Z. AOB: . ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ WW, 13-JAN-2002 15:03:34 _________________________________:_____________________________________ Wilfried Woeber : e-mail: Woeber(a)CC.UniVie.ac.at UniVie Computer Center - ACOnet : Tel: +43 1 4277 - 140 33 Universitaetsstrasse 7 : Fax: +43 1 4277 - 9 140 A-1010 Vienna, Austria, Europe : RIPE-DB: WW144, PGP keyID 0xF0ACB369 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

1 0

Routing Registry BOF at Miami NANOG (Feb 11)
by Larry J. Blunk 11 Jan '02

11 Jan '02

Merit will be hosting a Routing Registry BOF at the next NANOG meeting in Miami on Feb 11. Details on the meeting can be found at http://www.nanog.org/mtg-0202/. We also welcome any presentation proposals from interested parties. Regards, Larry J. Blunk Lead Engineer - Merit RADB project

1 0

IRT object
by Andrei Robachevsky 11 Jan '02

11 Jan '02

Dear Colleagues, The irt object has been implemented and we put the new software on our test whois server (please see http://www.ripe.net/ripencc/pub-services/db/testdb.html). We also took liberty to make some modifications to the original proposal (http://www.ripe.net/ripe/mail-archives/db-wg/20010701-20020101/msg00036.html). Namely: - an "irt-nfy:" attribute is now optional/multiple/inevrse key in the irt object. It specifies an e-mail address used to send notifications whenever a reference to the object is added or removed. We thought it would be very useful from security prospective. - a name of an irt object should start with "IRT-" prefix reserved for this type of object. This will eliminate possible confusion, such as mixing up with mntner. We have created some object to play with: irt: IRT-TEST inetnum: 172.18.0.0 - 172.25.255.255 So you may try: $ whois -h test-whois IRT-TEST $ whois -h test-whois -i mnt-irt IRT-TEST $ whois -h test-whois -i irt-nfy bitbucket(a)ripe.net $ whois -h test-whois 172.18.10.1 $ whois -h test-whois -c 172.18.10.1 The irt objects are not created automatically, but are forwarded to RIPE Database Administration (ripe-dbm(a)ripe.net). But this restriction does not apply to the TEST database, so you may try to create one yourself. Have a nice weekend, Regards, Andrei Robachevsky DB Group Manager RIPE NCC

1 0

Status of IRT object
by Gabriella Paolini 07 Jan '02

07 Jan '02

Dear all, which is the status of IRT object? Is the trial implementation available? This example shows again the need for a specific object. If someone uses whois information from cyberabuse.org, as many times people does, that information is wrong, because ciberabuse guys used Admin-c email without considering all other informations in inetnum object and in role object. Thanks and regards, Gabriella Paolini INFN-GARR [ Informations about 193.204.17.211 ] IP range : 193.204.16.0 - 193.204.23.255 Network name : UNIBAS-NET AS number : AS137 Infos : Universita' Degli Studi Di Basilicata Infos : Centro Interfacolta' Servizi Informatici E Telematici - Cisit Phone number : +39 06 49914352 Country : IT Abuse contact : enzo.valente(a)infn.it Source : RIPE right information are inetnum: 193.204.16.0 - 193.204.23.255 netname: UNIBAS-NET descr: Universita' degli Studi di Basilicata descr: Centro Interfacolta' Servizi Informatici e Telematici - CISIT country: IT admin-c: EV182-RIPE tech-c: GL965-RIPE tech-c: GM976-RIPE ... remarks: To notify abuse mailto: cert(a)garr.it remarks: Multiple lan of university departments remarks: GARR - Italian academic and research network mnt-by: GARR-LIR ... role: GARR LIR ... trouble: To notify abuse mailto: cert(a)garr.it trouble: Information at http://www.lir.garr.it/ ... person: Enzo Valente address: INFN - Dipartimento di Fisica address: Universita' La Sapienza address: Piazzale Aldo Moro,2 address: I-00185 Roma address: Italy phone: +39 06 49914352 fax-no: +39 06 4957697 e-mail: Enzo.Valente(a)infn.it ...

4 3

RIPE robot response
by Antonia López 04 Jan '02

04 Jan '02

Dear all, Every time I update or add an object to the RIPE database, the robot give back an e-mail like that (although I send my e-mail with non formatted text). <<FAILED: >> The action is made, then why this e-mail?? Thanks. Antonia Lopez Serrano

3 2

RE: RIPE robot response
by Boyan Krosnov 04 Jan '02

04 Jan '02

You sent a mime message with two parts part one is text/html the other is text/plain. The robot cannot parse html so it sees it as plain text and tries to process it. It then fails, as could be expected. Then it processes the text/plain part of your message, which succeedes. Never send html formated messages to the robot or it will spit errors at you. If you are using a MS exchange server you can make it send messages to all addresses @ripe.net in plain text format. Drop me a private e-mail for the exact configuration. Regards, CCNP Boyan Krosnov Network Administrator Lirex Net phone: +359-2-91815 -----Original Message----- From: Antonia Lopez [mailto:antonia.lopez@abrared.com] Sent: Friday, January 04, 2002 6:18 PM To: 'db-wg(a)ripe.net' Subject: RIPE robot response Dear all, Every time I update or add an object to the RIPE database, the robot give back an e-mail like that (although I send my e-mail with non formatted text). <<FAILED: >> The action is made, then why this e-mail?? Thanks. Antonia Lopez Serrano

1 0

Re: RIPE robot response
by Janos Zsako 04 Jan '02

04 Jan '02

> From owner-db-wg(a)ripe.net Fri Jan 4 16:35:03 2002 > From: =?iso-8859-1?Q?Antonia_L=F3pez?= <antonia.lopez(a)abrared.com> Dear Antonia, This is probably because your mailer is cofigured to send the message BOTH in plaintext AND HTML (see below). The plaintext is interpreted correctly by the robot, while the HTML text is flaged with errors. Also, please note that you should change your password, as you forwarded it in cleartext to the list... :(( Best regards, Janos > Content-Type: multipart/alternative; > Content-Type: text/plain; > Dear all, > > Every time I update or add an object to the RIPE database, the robot give > back an e-mail like that (although I send my e-mail with non formatted > text). > > <<FAILED: >> > > The action is made, then why this e-mail?? > Content-Type: text/html; > Dear all, > > > Every time I update or add an object = > to the RIPE database, the robot give back an e-mail like that (although = > I send my e-mail with non formatted text).

1 0

RE: route object check-up
by Wilfried Woeber, UniVie/ACOnet 28 Dec '01

28 Dec '01

>while updating our database we've noticed some >inconsistency in ripe-db route object: 80.82.178.0/20 >so does anybody know why net and prefix combination for >route object could be invalid ? When the starting address is not properly aligned on a power of 2 for the block-size. The hint you are receiving, as attached below, is an indication for this, i.e. - 80.82.178.0 is only properly aligned for 2 consecutive Class Cs, 80.82.178.0 and 80.82.179.0), - whereas for a /20 (16 consecutive Class Cs) the proper alignment would be 80.82.176.0 Btw, the same issue might eventually arise when someone tries to do an aggregated announcement on the routing layer for this block. Possible solutions: - move the address block to the proper alignment - partiton the block into a set of smaller, consecutive, aligned chunks - try to register the address as a range (although this might trigger the same error message, depending on implementation of checks...) I guess the older versions of the software were less picky than the new implementation is... >With respect, >--- >RIPN Registry center >----- > >-------------------------------------------------------------------- > >Update [FAILED] > >dbupdate: >80.82.178.0/20 is an invalid net and prefix combination:- >length 20 (mask 255.255.240.0) is outside the range of 80.82.178.0 >Possible values with these combinates are: >based on length 80.82.176.0/20 >based on prefix 80.82.178.0/23 Regards, _________________________________:_____________________________________ Wilfried Woeber : e-mail: Woeber(a)CC.UniVie.ac.at UniVie Computer Center - ACOnet : Tel: +43 1 4277 - 140 33 Universitaetsstrasse 7 : Fax: +43 1 4277 - 9 140 A-1010 Vienna, Austria, Europe : RIPE-DB: WW144, PGP keyID 0xF0ACB369 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

3 2