- db-wg - mailman.ripe.net

Ref: Agenda points for the db-wg session (call for...)
by Wilfried Woeber, UniVie/ACOnet 17 Jan '03

17 Jan '03

Dear DB-WG! The RIPE NCC has already submitted items for the Jan 2003 DB-WG meeting's draft agenda. If anyone wants to propose issues/topics/items to be included, please sumbit them _asap_. I'll review the list of open actions and the minutes to complement the set of items from my end during the weekend. Regards, Wilfried. _________________________________:_____________________________________ 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

Updated proposal for clean-up of references by name
by Engin Gunduz 16 Jan '03

16 Jan '03

[Apologies for duplicate messages] Dear Colleagues, I'm attaching the updated proposal for automatically cleaning up the references by name in RIPE Whois Database. We have incorporated Janos's suggestions to the proposal. If no more comments come, we plan to apply the procedure described in the proposal in the first half of January, 2003. The operation should not take more than a day. We will notify the community about the actual date of the operation. Regards, Engin Gunduz ____________________________ RIPE NCC Database Group -------------- Proposed solution for cleaning up references by name and other invalid references in RIPE Whois Database Motivation: References by name and invalid references cause two main problems: 1. One reference by name in a single object locks all person and role objects with that name, that is, they cannot be deleted, because of referential integrity checks. 2. Having anything other than a NIC handle as a reference makes the implementation of whois database software considerably more complex, since the software needs to deal with these exceptions. This increases the coding time, maintenance time and testing time of the software. Classification of the inconsistencies we need to solve and proposed solutions: 1. Objects that refer to a person or role object by name. a. There is only one object with this name. 1. The referring inconsistent object is not maintained, or the maintainers of referring inconsistent object and the person/role object with this name are the same. Solution: Update the referring inconsistent object so that it will contain the NIC handle instead of the name. Add appropriate remarks and changed attributes to the object to explain the reason for update. 2. The referring object is maintained, and the maintainers are different from the maintainers of the object with the referred name (If the latter object is not maintained, then the maintainers are by definition different.) Solution: Create a role object with this name. It will list the role or person object with this name in its admin-c and tech-c attributes. Update the inconsistent object to refer to the NIC handle of this new role object. Add appropriate remarks and changed attributes to the object to explain the reason for update. b. There is no person or role object with this name: Solution: Create a person object with this name. Clearly mark this new object putting appropriate remarks attributes so that users will see it is actually a dummy object. Update the inconsistent object to refer to the NIC handle of this new person object. Add appropriate remarks and changed attributes to the object to explain the reason for update. Protect it with the inconsistent object's maintainer(s). c. There are multiple person and role objects with this name. Solution: Create a role object with this name. It will list all the other role and person objects with the same name in its admin-c and tech-c attributes. Update the inconsistent object to refer to the NIC handle of this new role object. Add appropriate remarks and changed attributes to the object to explain the reason for update. 2. Objects that refer to a non-existent NIC handle. Solution: Create a person object with that NIC handle. Clearly mark this new object so that users will see it is actually a dummy object. Name it "person: Place Holder Object". Protect it with the inconsistent object's maintainer(s). Note that there is no need to update the inconsistent object itself. 3. Objects that refer to a string which is neither a name, nor a NIC handle. For example, it might be a phone number in admin-c attribute, or it might be 'Gunduz', a string that can't be a NIC handle, as it's longer than 4 letters, nor can it be a name as it has only one word. Another example could be "Mr. Gunduz", which enters this category because "Mr" can't appear in a name of person/role object. Solution: Create a person object for each such reference. Name the object "person: Place Holder Object" and list the object that refers to it in its remarks attribute. Protect it with inconsistent object's maintainer(s). Then update the inconsistent object to refer to the NIC handle of this new place holder person object. In each case an object is updated, or created, send appropriate notifications (determined by "mnt-by" and "notify" attributes, as with all other updates). Please note that this proposal does not actually solve the problem of invalid contact information-- rather, it makes the data set more uniform, thus decreases the administration and development time of the whois database.

1 2

RE: [lir-wg] should aut-num object for IPv4 & IPv6 be the same?
by adrian.pauling＠bt.com 15 Jan '03

15 Jan '03

we have 'inet6num' objects separate to 'inetnum' to distinguish IPv6 address assignments from IPv4. Is the question should different 'peering-set' 'route' objects etc reflect different peerings for IPv6 from IPv4 that an assigned AS can have? This scenario is real as I can easily give a current (enterprise LIR) example. Regards, Adrian F Pauling BT Ignite Solutions

2 1

Implementing additional IPV4 Table into RIPEDB-Software (local use only)
by Arnd Vehling 08 Jan '03

08 Jan '03

Hello, i need to introduce a new IP V4 Table into the current release of the ripe-db software for a local project at Cable&Wireles. The best starting point for impletenting this is probably the definition of the current ipv4 definition of "inetnum" in attributes.xml. Befor i go ahead and start reading through tons of sourcecode, especially the one used for radix-tree operations, i thought i might ask here if someone has any general tips on what to change/not to change. Comment on the current definition if the inetnum object would also be of great help to me. I.e. which attributes to copy and which need to be changed etc. best regards, Arnd Vehling -- NetHead Network Design and Security Arnd Vehling av(a)nethead.De Gummersbacherstr. 27 Phone: +49 221 8809210 50679 Köln Fax : +49 221 8809212

1 0

ANNOUNCEMENT: ROUTING REGISTRY TRAINING COURSES
by RIPE NCC Training 20 Dec '02

20 Dec '02

Dear colleagues, [apologies for duplicate postings] As a new service to its members the RIPE NCC offers the "Routing Registry Training Courses". The main objective of the "Routing Registry Training Course" is to familiarise the LIRs with the features of the Routing Registry, while introducing the relevant tools and services of the RIPE NCC. The course also offers practical exercises. OVERVIEW OF TOPICS COVERED IN THE COURSE: - Routing Registry as a part of the RIPE Whois Database and the Internet Routing Registry (IRR) - Specifying routing policies using RPSL (Routing Policy Specification Language) and the RtConfig tool - IRRToolSet - Routing Information Service (RIS) - Routing Registry Consistency Check Project Please note that this is an advanced course, it will: - NOT teach the basics of routing. - NOT describe how to receive Internet resources from the RIPE NCC, nor how to operate a Local Internet Registry (LIR) The target audience of the course are technical staff of LIRs: e.g. network operators, engineers etc. This course is not intended for administrative or management staff (e.g. hostmasters) . It is assumed that all attendees will be familiar with BGP routing and with the RIPE DB (creating/modifying objects). This course is based on interactive exercises in a local network set-up. Please bring your laptops equipped with: - Working Ethernet card, - 3 meters of Ethernet cable, - Secure connection client, - Web browser. Please be prepared to share your laptop (or at least allow another attendee to "look over your shoulder") if necessary. The Routing Registry Training Course is conducted in the English language and is free of charge, since it is covered by the membership contribution. More information about the Routing Registry Training Courses can be found at: http://www.ripe.net/training/rr/ REGISTRATION: You can register for a course at the following URL: http://www.ripe.net/cgi-bin/trainingform.pl.cgi Or by completing the registration form at the end of this email and replying to <training(a)ripe.net> . In order to register for a Routing Registry Training Course you must be an employee of an LIR and either : - be an LIR contact - be confirmed by an LIR contact. LIR contacts are those employees of an LIR who are registered with RIPE NCC as authoritative contact persons. It is expected that most of those interested in the "Routing Registry Training Course" will not be an authoritative contact persons for their LIR, therefore will be refused by the course registration "robot". In order to be admitted to the course, confirmation email need to be send to <training(a)ripe.net> . Please approach the LIR contacts from your organisation personally, since the identity of LIR contacts is confidential, and the RIPE NCC is unable to divulge contact persons for any given LIR. Kind Regards, The RIPE NCC Training Team ----------------------------------------------------------------------- COURSE DATES AND VENUES ======================== Date: Friday 24th Janaury 2003 Time: 0900 - 1700 Location: Amsterdam, Netherlands AND: Date: Wednesday 12th March 2003 Time: 0900 - 1700 Location: London, United Kingdom ------------------------------------------------------------------------ REGISTRATION FORM ================= %START PART 1 - Registration 1) Your name Enter First name, Last name in full e.g. John Doe Mary-Beth Walton # NAME [ ] 2) Your Registry ID (format: country-code.<name of registry>) # REG [ ] 3) Your e-mail address # EMAIL [ ] 4) Your NIC handle (optional) # NICHANDLE [ ] 5) The Course you plan to attend (date and location) # COURSE [ ] %END

2 2

Proposal: key-cert update procedure change
by Dmitry Morozovsky 17 Dec '02

17 Dec '02

Dear colleagues, changing today our key-cert objects, I'd encounter a bit of inconvenience: one can't cut'n'paste key-cert object from whois request: [s]he should manually delete method/owner/fingerpr attributes. My suggestion is: if one of r/o attributes exists in the object processed, DB should check whether auto-computed value of the attribute matches supplied value; - if yes, no error/warning should be produced - if no -- it should of course be reported; I think, it's up to the community to decide whether this condition should be treated as error or simply as warning. I think, the same procedure should be also applied to other de-facto read-only attributes, such as person name in person object. Sincerely, D.Marck [DM5020, DM268-RIPE, DM3-RIPN] ------------------------------------------------------------------------ *** Dmitry Morozovsky --- D.Marck --- Wild Woozle --- marck(a)rinet.ru *** ------------------------------------------------------------------------

2 2

Re: [db-wg] Proposal for automated clean-up of references by name
by Janos Zsako 05 Dec '02

05 Dec '02

> From db-wg-admin(a)ripe.net Thu Dec 5 11:08:46 2002 > From: Gert Doering <gert(a)space.net> > On Thu, Dec 05, 2002 at 10:42:53AM +0100, Engin Gunduz wrote: > > We have prepared a proposal to automatically clean-up references > > by name and other invalid references in RIPE Whois Database. > > I like the proposal. Go ahead! Yes, I agree too. > (Yes, it will potentially create lots of "ugly looking" objects - but > the situation right now, where you're not able to delete person objects > because they are locked by a reference-by-name, is much worse) I have one question though: who will be the maintainer of the new "ugly" object created? I have the feeling it could/should be the maintainer of the object that will reference it. Janos

1 0

Proposal for automated clean-up of references by name
by Engin Gunduz 05 Dec '02

05 Dec '02

[Apologies for duplicate messages] Dear Colleagues, We have prepared a proposal to automatically clean-up references by name and other invalid references in RIPE Whois Database. The proposal aims at fixing objects like (as an hypothetical example): inetnum: 58.210.0.0 - 58.210.255.255 netname: EXAMPLE-OBJECT descr: Example inconsistent object country: EU admin-c: Engin Gunduz tech-c: Engin Gunduz status: ASSIGNED PA mnt-by: RIPE-NCC-NONE-MNT changed: engin(a)ripe.net 19940508 source: RIPE Note that in admin-c and tech-c attributes there are names, instead of NIC handles, which is currently regarded as a syntax error. Such objects were introduced into RIPE Whois Database when the syntax checks were not as strict as today. There are 2114 (out of approximately 2 million references) such invalid references as of December 2nd, 2002. The proposal will affect only the objects that have invalid references where a NIC handle should be used. Please see http://www.ripe.net/db/dbconstat/ for more information on Database Consistency and Statistics Project which deals with this and other types of inconsistencies in RIPE Database. Please find our proposal below. We would like to discuss the proposal until December 20th, 2002, at which date we will sum up the discussion, and perhaps begin to implement the resulting solution. Regards, Engin Gunduz RIPE NCC Database Group -------------- Proposed solution for cleaning up references by name and other invalid references in RIPE Whois Database Motivation: References by name and invalid references cause two main problems: 1. One reference by name in a single object locks all person and role objects with that name, that is, they cannot be deleted, because of referential integrity checks. 2. Having anything other than a NIC handle as a reference makes the implementation of whois database software considerably more complex, since the software needs to deal with these exceptions. This increases the coding time, maintenance time and testing time of the software. Classification of the inconsistencies we need to solve and proposed solutions: 1. Objects that refer to a person or role object by name. a. There is only one object with this name. Solution: Update the inconsistent object so that it will contain the NIC handle instead of the name. Add appropriate remarks and changed attributes to the object to explain the reason for update. b. There is no person or role object with this name: Solution: Create a person object with this name. Clearly mark this new object putting appropriate remarks attributes so that users will see it is actually a dummy object. Update the inconsistent object to refer to the NIC handle of this new person object. Add appropriate remarks and changed attributes to the object to explain the reason for update. c. There are multiple person and role objects with this name. Solution: Create a role object with this name. It will list all the other role and person objects with the same name in its admin-c and tech-c attributes. Update the inconsistent object to refer to the NIC handle of this new role object. Add appropriate remarks and changed attributes to the object to explain the reason for update. 2. Objects that refer to a non-existent NIC handle. Solution: Create a person object with that NIC handle. Clearly mark this new object so that users will see it is actually a dummy object. Name it "person: Place Holder Object". Note that there is no need to update the inconsistent object itself. 3. Objects that refer to a string which is neither a name, nor a NIC handle. For example, it might be a phone number in admin-c attribute, or it might be 'Gunduz', a string that can't be a NIC handle, as it's longer than 4 letters, nor can it be a name as it has only one word. Another example could be "Mr. Gunduz", which enters this category because "Mr" can't appear in a name of person/role object. Solution: Create a person object for each such reference. Name the object "person: Place Holder Object" and list the object that refers to it in its remarks attribute. Then update the inconsistent object to refer to the NIC handle of this new place holder person object. In each case an object is updated, send appropriate notifications (determined by "mnt-by" and "notify" attributes, as with all other updates). Please note that this proposal does not actually solve the problem of invalid contact information-- rather, it makes the data set more uniform, thus decreases the administration and development time of the whois database.

2 1

Re: [lir-wg] Re: ERX of v4 address space proposal
by Andrei Robachevsky 20 Nov '02

20 Nov '02

Peter Galbavy wrote: > 192.0.0.0/8 ? > > The swamp is coming to RIPE or which poor RIR ? Our current data shows that around 753 networks will go to APNIC, 9072 stay with ARIN, 92 will go to LACNIC and 3176 to RIPE NCC. 192/8 will be processed as last and sweet piece, and statistics will be updated. > > Peter Regards, Andrei > > ----- Original Message ----- > From: "Andrei Robachevsky" <andrei(a)ripe.net> > To: <db-wg(a)ripe.net>; <lir-wg(a)ripe.net> > Sent: Wednesday, November 20, 2002 1:54 PM > Subject: [lir-wg] Re: ERX of v4 address space proposal > > > >>Dear Colleagues, >> >>You may find a list of networks (/8's) containing allocations that will >>be transferred by the ERX project at the following URL: >>http://www.ripe.net/db/erx/erx-ip/ >> >>A detailed list of which networks will be transferred, and to which RIR, >>will be posted before each /8 is transferred. A full schedule for each >>network will also be published for each /8. >> >>Regards, >> >>Andrei Robachevsky >>CTO, RIPE NCC >> >> >

1 0

ERX of v4 address space proposal
by Andrei Robachevsky 20 Nov '02

20 Nov '02

Dear Colleagues, Please find attached the proposal for the ERX of IPv4 address space for your review. This proposal was reviewed by the ERX-Taskforce that was set-up at the RIPE 43 meeting. Comments received were summarised and sent to both Database and LIR Working Group mailing lists. We would appreciate your feedback on this proposal no later than Wednesday, 13 November 2002. A timeline for this project will be issued along with final project description. Regards, Andrei Robachevsky RIPE NCC ERX of v4 address space ----------------------- INTRODUCTION When ARIN began operations, it inherited classful blocks of addresses ("legacy" space) that were issued to organisations not in ARIN's region. The Regional Internet Registries (RIRs) are co-ordinating the transfer of these records from the ARIN Database to RIPE Database or APNIC Database, as appropriate. This is based on geographic location of each resource holder. The affected records include class B networks and class C ranges, particularly those issued from 192/8. ARIN also maintains reverse delegation for all legacy address space. Once the transfer is complete, this responsibility will be distributed among the RIRs according to their respective regions. BENEFITS This will enable resource holders to maintain all their records in one database and End Users to interface with just one RIR for all database and reverse delegation matters. This effort will also help to locate address holders and recover unused or underutilised address space. IMPACT ON RESOURCE HOLDERS Entities holding legacy space to be transferred can expect the process to be largely transparent. While it will not affect network status in any way, resource holders might be required to create contact identifiers with the proper registry if they do not already exist. Queries made in WHOIS for an IP block that has been transferred to another RIR will be directed to the appropriate RIR. APPROACH The plan is to perform the transfer by /8. For each /8 the following tasks have to be performed: 1. Conflicts (contacts and description) to be resolved 2. Records and associated documentation to be transferred 3. Reverse delegation to be set up A total of 47 /8's with 8030 records have to be transferred. CONFLICTS The following types of conflicts have been identified: C1. Record exists in ARIN DB only. There is no exact matching (range wise) record in the RIPE DB. Proposal: to update internal documentation, create the record in the RIPE DB and protect with a unique generated mntner. C2. Range matches records in both ARIN and RIPE DBs. Meaning that contact and description may be different. Most cases indicate out of date information in one of the Databases, not real conflicts or attempts to hijack address space. What happened in most cases is that people started maintaining their allocation or assignement in the RIPE DB, especially since RIPE DB started to support the Routing Registry. Proposal: Notify conflicts and give time to reach consensus. After the deadline merge those who haven't responded, but _do_not_lock_ (keep the same mntner). C3.0 Record exists in the RIPE DB only. C3.1. One reason why such situation may exist is that this is a valid RIPE NCC allocation. Proposal: To preserve information in the RIPE DB. C3.2 Another situation is that the registration data are simply garbage. Proposal: Notify, give time for explanations and clean up in the end. PROCEDURE for a /8 1. Pre transfer 1.1 Initial dump is prepared for transfer by ARIN 1.2 Announcement is sent to ARIN's contacts 1.3 Reverese delegation domain space is cleaned up in the RIPE DB (reverse domain objects for which no delegation was provided are deleted) 2. Transfer 2.1 Final dump is prepared by ARIN 2.2 C1 group: database records are imported, documentation is updated, contacts are notified 2.3 C2 group: contacts (ARIN+RIPE) are asked to reach consensus. 2.4 C3.2 group: contacts are notified of possible deletion. 2.5 DNS is updated: domain objects are generated, zone (full or partial) is generated. 3. Conflict resolution 3.1 C2 group: non responding records are merged but _not_locked_. 3.2 C3.2 group: records without good reasons are deleted. -------

4 7