- db-wg - mailman.ripe.net

Minutes for the RIPE 46 meeting Database Working Group Sessions
by Can Bican 12 Sep '03

12 Sep '03

Dear Colleagues, Here are the minutes for the two sessions of the Database Working Group in RIPE 46. Please let me know if they need additions or corrections. Best Regards, -- Can Bican DB Group RIPE NCC

1 1

Re: [db-wg] Re: NCC#2003094280 SUCCESS: Bulk deletion of /24 detail
by Janos Zsako 12 Sep '03

12 Sep '03

> From db-wg-admin(a)ripe.net Fri Sep 12 11:15:10 2003 Dear all, > Regarding the SUCCESS reply it was a decision that we made when we re-structured > dbupdate that we only reported FAILUE if we detected an error. If there is no > error we report SUCCESS. This means that SUCCESS is now returned for a blank > update message, a message where none of the objects are recognised (as in your > case) and when no operation is performed because the submitted object is > identical to the one currently in the database. > > It would not be difficult to change this behaviour if it is generally felt that > the current scheme is mis-leading. Perhaps we can have some opinions about this > from other users on the db-wg mailing list. As input for the WG, here is the relevant part of the private mail I sent to Niall: ----- Begin Included Message ----- My understanding is that the parser did not find any object, i.e. any task to be performed, therefore it did not flag any error, so the whole mail was processed with "sucess"... I admit this is definitely misleading, so it may be useful to treat this as a special case (i.e. the case of an empty mail that has a warning). [In case you submit several objects or tasks, one of which at least is correct and you also have ignored objects due to wrong syntax, the result SUCCESS is less misleading in my view, as at least the correct task was performed with success...] ----- End Included Message ----- In other words, do not flag with SUCCESS an empty message. I recommend to flag it with something like UNKNOWN. Janos

1 0

Re: NCC#2003092385 [db-wg] uodate failure
by Wilfried Woeber, UniVie/ACOnet 09 Sep '03

09 Sep '03

Hi Randy, thanks for bringing this up! [ the "keyword noise" had me confused too, a short while ago :-( ] I am wondering whether "inline" error messages are such a good idea? I guess there's a good chance that they remain in the updated and re-sent message. That could give us another set of failure modes. _I_ would probably go for a more descriptive error message at the end, including the _attribute_ that made the reference, _and_ the _string_ used as a reference. -WW ______________________________________________________________________ From: Randy Bush <randy(a)psg.com> To: RIPE Database Administration <ripe-dbm(a)ripe.net> CC: db-wg(a)ripe.net Subject: Re: NCC#2003092385 [db-wg] uodate failure Date: Mon, 8 Sep 2003 06:07:48 -0700 btw, to be very clear, and as a once-upon-a-time compiler hacker who would have been severely excoriated for poor diagnostic reporting, what i would have liked to have received would be something along the lines of Modify FAILED: [aut-num] AS3130 ***Info: Authorisation passed ***Info: Syntax check passed aut-num: AS3130 as-name: RGNET-3130 descr: RGnet/PSGnet admin-c: RB366 ^ unknown object referenced tech-c: RB366 ^ unknown object referenced import: from AS1239 accept ANY import: from AS2914 accept ANY import: from AS3927 accept AS3927 import: from AS3970 accept AS3970 import: from AS4128 accept AS4128 import: from AS101 accept AS-PNW-GIGAPOP import: from AS11726 accept AS-ACUMEN import: from AS12111 accept AS-OEN import: from AS16713 accept AS-NOANET-PEERING import: from AS2044 accept AS2044 import: from AS3491 accept AS-CAIS import: from AS4513 accept AS-GLOBIX import: from AS4565 accept AS-EPOCH import: from AS6456 accept AS6456 import: from AS7610 accept AS-SINGAREN import: from AS8075 accept AS8075 export: to AS1239 announce AS-RGNET export: to AS2914 announce AS-RGNET export: to AS3927 announce ANY export: to AS3970 announce ANY export: to AS4128 announce ANY export: to AS101 announce AS-RGNET export: to AS11726 announce AS-RGNET export: to AS12111 announce AS-RGNET export: to AS16713 announce AS-RGNET export: to AS2044 announce AS-RGNET export: to AS3491 announce AS-RGNET export: to AS4513 announce AS-RGNET export: to AS4565 announce AS-RGNET export: to AS6456 announce AS-RGNET export: to AS7610 announce AS-RGNET export: to AS8075 announce AS-RGNET notify: rw(a)rg.net mnt-by: MAINT-RGNET changed: randy(a)psg.com 20030905 source: RIPE ***Info: Authorisation for [aut-num] AS3130 using mnt-by: authenticated by: MAINT-RGNET randy --------------------------------------------------------------------------------

4 3

Re: NCC#2003092385 [db-wg] uodate failure
by Denis Walker 08 Sep '03

08 Sep '03

"Wilfried Woeber, UniVie/ACOnet" <woeber(a)cc.univie.ac.at> writes: * Hi Randy, * * thanks for bringing this up! * [ the "keyword noise" had me confused too, a short while ago :-( ] I think we assumed that people would associate 'keywords' with those words in the subject line of a mail message or entered separately with the web updates form. Perhaps we need to add something to the message to make it more clear. * * I am wondering whether "inline" error messages are such a good idea? * I guess there's a good chance that they remain in the updated and * re-sent message. That could give us another set of failure modes. * * _I_ would probably go for a more descriptive error message at the end, * including the _attribute_ that made the reference, _and_ the _string_ * used as a reference. This is the way we have planned to do it and the code has already been written for the dbupdate part. But these error messages originate in the server back end code where all referential integrity checks are performed. It is on our 'ToDo' list to review this section of the code. cheers denis * * -WW * ______________________________________________________________________ * * From: Randy Bush <randy(a)psg.com> * To: RIPE Database Administration <ripe-dbm(a)ripe.net> * CC: db-wg(a)ripe.net * Subject: Re: NCC#2003092385 [db-wg] uodate failure * Date: Mon, 8 Sep 2003 06:07:48 -0700 * * btw, to be very clear, and as a once-upon-a-time compiler hacker * who would have been severely excoriated for poor diagnostic * reporting, what i would have liked to have received would be * something along the lines of * * Modify FAILED: [aut-num] AS3130 * ***Info: Authorisation passed * ***Info: Syntax check passed * * aut-num: AS3130 * as-name: RGNET-3130 * descr: RGnet/PSGnet * admin-c: RB366 * ^ unknown object referenced * tech-c: RB366 * ^ unknown object referenced * import: from AS1239 accept ANY * import: from AS2914 accept ANY * import: from AS3927 accept AS3927 * import: from AS3970 accept AS3970 * import: from AS4128 accept AS4128 * import: from AS101 accept AS-PNW-GIGAPOP * import: from AS11726 accept AS-ACUMEN * import: from AS12111 accept AS-OEN * import: from AS16713 accept AS-NOANET-PEERING * import: from AS2044 accept AS2044 * import: from AS3491 accept AS-CAIS * import: from AS4513 accept AS-GLOBIX * import: from AS4565 accept AS-EPOCH * import: from AS6456 accept AS6456 * import: from AS7610 accept AS-SINGAREN * import: from AS8075 accept AS8075 * export: to AS1239 announce AS-RGNET * export: to AS2914 announce AS-RGNET * export: to AS3927 announce ANY * export: to AS3970 announce ANY * export: to AS4128 announce ANY * export: to AS101 announce AS-RGNET * export: to AS11726 announce AS-RGNET * export: to AS12111 announce AS-RGNET * export: to AS16713 announce AS-RGNET * export: to AS2044 announce AS-RGNET * export: to AS3491 announce AS-RGNET * export: to AS4513 announce AS-RGNET * export: to AS4565 announce AS-RGNET * export: to AS6456 announce AS-RGNET * export: to AS7610 announce AS-RGNET * export: to AS8075 announce AS-RGNET * notify: rw(a)rg.net * mnt-by: MAINT-RGNET * changed: randy(a)psg.com 20030905 * source: RIPE * * ***Info: Authorisation for [aut-num] AS3130 * using mnt-by: * authenticated by: MAINT-RGNET * * randy * * ---------------------------------------------------------------------------- * ---- *

1 0

Re: [db-wg] Re: "changed:" attribute
by Denis Walker 08 Sep '03

08 Sep '03

Steve Atkins <steve(a)blighty.com> writes: * On Mon, Sep 08, 2003 at 03:21:20PM +0100, Rodney Tillotson wrote: * > 2. * > > Secondly, the query users do not really know what "changed:" attribute * > > means. ... most users and automated tools pick anything that resembles * > > an email address in the whois query result and send them abuse * > > complaints. * > * > The question here is about external presentation; who sees "changed:", * > what they are looking for when they see it, what they think it means * > and what they do with it. * > * > * > The anti-spam WG is preparing a broad proposal, but one early thought * > is to provide different database views for the different classes of * > user. Casual users looking for contact information in the default * > output are an important class, probably not the most well-informed or * > patient, and it may be worth suppressing "changed:" in the view they * > get. This applies whatever decisions we make about the internal * > behaviour or the way LIRs will use the attribute. * * That seems like a good idea. Many casual users of the RIPE database do * not query it directly, rather through a small number of web-based * gateways. Would it be worth asking the owners of those gateways * to filter out the changed field now, to see what effect that has? But then someone will realise the changed lines are missing and think we are hiding something. So they will write a tool to show them and give them even more significance. cheers denis * * Cheers, * Steve *

2 1

The New "organisation object" Proposal
by DB-News 08 Sep '03

08 Sep '03

Dear Colleagues, [apologies for duplicate messages] Following discussions at the RIPE NCC about a new Whois Database object, the organisation object, we have modified the previous proposal that you can find at: http://www.ripe.net/ripe/mail-archives/db-wg/2003/msg00013.html The main changes from the previous proposal are: - a "country:" attribute is added. - "org-type:" and "org-name:" attributes are added, and "organisation:" attribute is now a NIC handle, which will be automatically assigned during the creation of the object. The motivation for this is explained in Section 3 of the proposal. - Hierarchical names can be removed, mainly for simplicity. - Differing query results on the whois server introduced in section 5. According to this, the relevant organisation objects are appended to the query results (because the organisation that holds the resource is at least as important as the person/role objects mentioned in the resource object). Please let us know your comments and concerns about the new proposal attached to this mail. Thanks for your time and best regards, -- Engin Gunduz RIPE NCC Database Group An organisation object in the RIPE Database -------------------------------------------- 1. Motivation ------------------- Currently the RIPE Database stores two main types of contact information: person and role objects. The person and role objects provide a way to contact people responsible for operations or usage of the resources represented in the RIPE Database (IP blocks, autonomous systems, and domain names). However, none of these provide an easy way of mapping resources to a particular organisation. A user must first find an object containing contact information for that organisation. Then, assuming all of the organisation's objects refer to this contact information, the user must perform an inverse query to obtain a list of objects referencing the specified person or role. This indirect process can be somewhat obscure and therefore a request for a more direct way of attaching an object to an organisation is seen as a useful addition to the RIPE Database. This document is a proposal for an organisation object in the RIPE Database and the necessary database functionality. 2. The organisation object ----------------------- The organisation object provides information identifying an organisation such as a company, charity or university, that is a holder of a network resource whose data is stored in the RIPE Database. The organisation object is identified by a unique ID specified in the "organisation:" attribute which is the primary key. An organisation object can be referenced from other types of objects using an "org:" attribute. All objects associated with a particular organisation can be retrieved by performing an "inverse query" for the organisation's ID (handle) used in the "org:" attribute of database objects. Following is a template for the proposed organisation object and an example. All attributes except the "ref-nfy:", "organisation:", "org-name:" and "org-type:" have their usual meanings. organisation: [mandatory] [single] [primary/look-up key] org-name: [mandatory] [single] [look-up key] org-type: [mandatory] [single] [ ] descr: [mandatory] [multiple] [ ] remarks: [optional] [multiple] [ ] address: [mandatory] [multiple] [ ] country: [mandatory] [single] [ ] phone: [optional] [multiple] [ ] fax-no: [optional] [multiple] [ ] e-mail: [mandatory] [multiple] [look-up key] admin-c: [mandatory] [multiple] [inverse key] tech-c: [mandatory] [multiple] [inverse key] ref-nfy: [optional] [multiple] [inverse key] notify: [optional] [multiple] [inverse key] mnt-by: [mandatory] [multiple] [inverse key] changed: [mandatory] [multiple] [ ] source: [mandatory] [single] [ ] 3. New attributes ---------------------- "organisation:" Specifies the ID of an organisation object. An organisation ID is made up of 'ORG-' prefix, followed by 2 to 4 letters, digits, a dash and is followed by the database source (in the RIPE Whois Database this is 'RIPE'). For example: ORG-RT34-RIPE Note that all parts are mandatory, thus ORG-RT-RIPE would be an invalid ID as it is missing the numeric part. Organisation object IDs are auto-generated similar to the way person/role "nic-hdl:" attributes are auto-generated. The user has to specify the ID of an organisation object as ORG-AUTO-<digit> during creation of the object, then it will be assigned an appropriate ID. The organisation ID is assigned using the "org-name:" attribute of the object. The user can specify the letter combination he/she prefers. For example if the user wants TTR as the letter combination, in the organisation ID, then ORG-AUTO-1TTR should be put into "organisation:" attribute during the creation of the object. The organisation ID cannot be reused. If an organisation ID was used in the past by an organisation object and then deleted, this ID cannot be used in new organisation objects. The auto-generation of organisation IDs and preventing reuse of them simplifies the external references. Note that when an organisation changes name, the "org-name:" can be modified accordingly. There is no need to change the organisation ID. "org-name:" Specifies the name of the organisation that this organisation object represents in the whois database. This is an ASCII-only text attribute. The restriction is because this attribute is a look-up key and the whois protocol does not allow specifying character sets in queries. The user can put the name of the organisation in non-ASCII character sets in the "descr:" attribute if required. "org-type:" Specifies the type of the organisation. The possible values are 'IANA' for Internet Assigned Numbers Authority, 'RIR' for Regional Internet Registries, 'NIR' for National Internet Registries, 'LIR' for Local Internet Registries, and 'NON-REGISTRY' for all other organisations. "ref-nfy:" Specifies the e-mail address to be notified when a reference to the organisation object is added or removed. An e-mail address as defined in RFC 2822. [This functionality is used in the "irt-nfy:". The name of this attribute is more neutral allowing reuse in other object types.] "country:" Specifies the two-letter ISO3166 country code of the country where this organisation resides. "address:" Specifies the address of the organisation. This is a free-text attribute. "org:" May be included in any other object type. It points to an existing organisation object representing the entity that holds the resource. The value of this attribute is the ID of the organisation object. It is mandatory in the aut-num objects, and inetnum and inet6num objects with "ALLOCATED-BY-IANA", "ALLOCATED-BY-RIR", "ALLOCATED-BY-RIR NON-PORTABLE", "ALLOCATED-BY-RIR PORTABLE" and "ALLOCATED-BY-RIR UNSPECIFIED" values. It is optional in all other objects, and it is single valued in all objects. 4. Authorisation checks ---------------------------------- When modifying an organisation object the update must pass authorisation checks specified by one of the mntners listed in the "mnt-by:" attributes of the organisation object. When adding an "org:" attribute to an object, the update of the object should pass the following authorisation checks: - from one of the maintainers of the organisation object - from one of the maintainers of the object being updated 5. Query changes ---------------- If a whois query returns an object with an "org:" attribute, the organisation object mentioned in this attribute is also appended to the query results. This behaviour can be disabled by using the '-r' flag in the query. 6. Examples ------------ A basic organisation object: organisation: ORG-RSIS54-RIPE org-name: Random Street Internet Services org-type: LIR descr: An example organisation address: Random St. city: Amsterdam country: NL phone: +31 123 4567 fax-no: +31 123 4568 e-mail: contact(a)example-org.net admin-c: EXAM1-RIPE tech-c: EXAM2-RIPE notify: ripe-mailbox(a)example-org.net ref-nfy: ripe-mailbox(a)example-org.net mnt-by: EXAMPLE-MNT changed: someguy(a)example-org.net 20030121 source: RIPE A network that references an organisation object: inetnum: 192.168.86.0 - 192.168.86.255 netname: EXAMPLE-NET-86 descr: Sample network org: ORG-RSIS54-RIPE country: NL admin-c: JE1-RIPE tech-c: JE2-RIPE status: ALLOCATED-BY-RIR PORTABLE mnt-by: EXAMPLE-MNT mnt-lower: EXAMPLE-MNT changed: someguy(a)example-org.net 20030122 source: RIPE A query for this inetnum object: % whois 192.168.86.251 inetnum: 192.168.86.0 - 192.168.86.255 netname: EXAMPLE-NET-86 descr: Sample network org: ORG-RSIS54-RIPE country: NL admin-c: JE1-RIPE tech-c: JE2-RIPE [...] source: RIPE organisation: ORG-RSIS54-RIPE org-name: Random Street Internet Services org-type: LIR descr: An example organisation address: Random St. address: The Netherlands phone: +31 123 4567 fax-no: +31 123 4568 e-mail: contact(a)example-org.net admin-c: EXAM1-RIPE tech-c: EXAM2-RIPE [...] source: RIPE person: John Example nic-hdl: JE1-RIPE [...] source: RIPE person: John Example Jr nic-hdl: JE2-RIPE [...] source: RIPE

17 35

Re: NCC#2003092385 [db-wg] uodate failure
by RIPE Database Administration 08 Sep '03

08 Sep '03

Dear Randy, I've sent the answer in the morning, strangely it didn't reach you. Please find it below. If you have any more questions, please contact <ripe-dbm(a)ripe.net>. Regards, Katie Petrusha ____________________________ RIPE Database Administration. From: RIPE Database Administration <ripe-dbm(a)ripe.net> To: Randy Bush <randy(a)psg.com> FCC: cur Subject: Re: NCC#2003091799 FAILED: as3130 Reply-To: ripe-dbm(a)ripe.net Date: Mon, 08 Sep 2003 10:43:51 +0200 Dear Randy, At the moment person/role object with nic-handle RB366 doesn't exist in the RIPEDatabase, therefore you can't reference it from any other object. However RB366-ARIN exists and seems to be yours, maybe you meant to use that one? If you have any more questions, please contact <ripe-dbm(a)ripe.net>. Regards, Katie Petrusha ____________________________ RIPE Database Administration. Original message follows: ------------------------ no answer from ripe-dbm(a)ripe.net, so figured to ask here. what did i do wrongly? (original was pgp wrapped) To: randy(a)psg.com Subject: FAILED: as3130 Date: Fri, 5 Sep 2003 23:08:42 +0200 (MEST) > From: Randy Bush <randy(a)psg.com> > Subject: as3130 > Date: Fri, 5 Sep 2003 14:08:25 -0700 > Reply-To: randy(a)psg.com > Message-ID: <E19vNog-0004bn-CN(a)ran.psg.com> SUMMARY OF UPDATE: Number of objects found: 1 Number of objects processed successfully: 0 Create: 0 Modify: 0 Delete: 0 No Operation: 0 Number of objects processed with errors: 1 Create: 0 Modify: 1 Delete: 0 Syntax Errors: 0 DETAILED EXPLANATION: ***Warning: Invalid keyword(s) found: as3130 ***Warning: All keywords were ignored ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ The following object(s) were found to have ERRORS: --- Modify FAILED: [aut-num] AS3130 ***Info: Authorisation passed ***Info: Syntax check passed aut-num: AS3130 as-name: RGNET-3130 descr: RGnet/PSGnet admin-c: RB366 tech-c: RB366 import: from AS1239 accept ANY import: from AS2914 accept ANY import: from AS3927 accept AS3927 import: from AS3970 accept AS3970 import: from AS4128 accept AS4128 import: from AS101 accept AS-PNW-GIGAPOP import: from AS11726 accept AS-ACUMEN import: from AS12111 accept AS-OEN import: from AS16713 accept AS-NOANET-PEERING import: from AS2044 accept AS2044 import: from AS3491 accept AS-CAIS import: from AS4513 accept AS-GLOBIX import: from AS4565 accept AS-EPOCH import: from AS6456 accept AS6456 import: from AS7610 accept AS-SINGAREN import: from AS8075 accept AS8075 export: to AS1239 announce AS-RGNET export: to AS2914 announce AS-RGNET export: to AS3927 announce ANY export: to AS3970 announce ANY export: to AS4128 announce ANY export: to AS101 announce AS-RGNET export: to AS11726 announce AS-RGNET export: to AS12111 announce AS-RGNET export: to AS16713 announce AS-RGNET export: to AS2044 announce AS-RGNET export: to AS3491 announce AS-RGNET export: to AS4513 announce AS-RGNET export: to AS4565 announce AS-RGNET export: to AS6456 announce AS-RGNET export: to AS7610 announce AS-RGNET export: to AS8075 announce AS-RGNET notify: rw(a)rg.net mnt-by: MAINT-RGNET changed: randy(a)psg.com 20030905 source: RIPE ***Error: Unknown object referenced ***Info: Authorisation for [aut-num] AS3130 using mnt-by: authenticated by: MAINT-RGNET ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For assistance or clarification please contact: RIPE Database Administration <ripe-dbm(a)ripe.net>

6 7

Suggestion for more user-friendly auto-dbm behaviour
by Niall O'Reilly 08 Sep '03

08 Sep '03

After a long period of not dealing with the RIPE database, I'm getting around to some overdue housekeeping. I've been surprised at how unfriendly the interface is. The following suggestions arise from this experience. First suggestion ... > ***Warning: Invalid keyword(s) found: Update > ***Warning: All keywords were ignored It could be more clearly put, that this refers only to the 'Subject' line. Why not replace the text, "All keywords were ignored" with "Entire subject line was ignored". Second suggestion ... > ***Error: Required attribute "netname" is missing > ***Error: Required attribute "country" is missing > ***Error: Required attribute "status" is missing > ***Error: Required attribute "mnt-by" is missing I didn't wish to modify any of these just now. Why not proceed as follows: 1. Load (partial) object from transaction. 2. Retrieve matching object (if any) from database. 3. Perform go/no-go tests (authorization, validity ...) 4. Merge transaction with what was retrieved (may be null). 5. Perform any completeness/self-consistency tests. 6. Commit update unless errors were found. 7. Notify parties involved. That should be worth at least EUR 0.02! I'll send on the invoice. 8-) Niall O'Reilly

1 0

summary of organisation object discussions
by Engin Gunduz 08 Sep '03

08 Sep '03

Dear Colleagues, [apologies for duplicate messages, and for having misspelled ncc-services-wg(a)ripe.net in the previous posting] I'd like to summarise the discussions took place in the last two weeks in the working group mailing lists after the publication of the organisation object proposal ( http://www.ripe.net/ripe/mail-archives/db-wg/2003/msg00159.html ) and after the presentation in the DB-WG session last week in RIPE 46 Meeting ( http://www.ripe.net/ripe/meetings/ripe-46/presentations/ripe46-db-organisat… ). 1. "org:" attribute mandatory or optional in aut-num objects? The proposal mentioned above specified "org:" attribute in aut-num objects to be mandatory. However, often LIRs request an ASN for their customers and the RIPE NCC does not have enough information for them to fill in an organisation object completely. Rather than putting invalid information on the whois database, the RIPE NCC thinks it is better to make "org:" attribute optional in aut-nums. 2. "org:" attribute multi- or single-valued? The proposal specified "org:" attribute single-valued in all objects. During the discussions in the DB WG, commentors from the audience stated that in some cases a person object works for more than one companies, thus one might want to put multiple "org:" attributes in his/her person object. Similar reasoning could apply to other object types as well. However, the "org:" object specifies the holder of an Internet resource in inetnum, inet6num and aut-num objects, thus it must be single-valued in these objects. For the rest of the objects, we propose to modify the rule so it can be multi-valued. 3. addition of "mnt-ref:" attribute. The proposal suggested the use of "mnt-by:" attribute for both authorisation of modification of the organisation object and the authorisation of adding a reference to the organisation object from other objects. This has a shortcoming when the organisation object is protected by an entity other than the organisation itself (for example, an LIR, whose organisation object is protected by the RIPE NCC). We would propose to introduce mandatory "mnt-ref:" attribute to control the references to the organisation object, while "mnt-by:" protects the organisation object itself, as usual. 4. "org:" attribute in the organisation object itself. The proposal suggested the use of "org:" attribute in non-organisation objects. I think it could be used in organisation objects as well, to indicate business relations like being a dependent company to another one. 5. "business-id:" attribute. It was suggested to add an attribute to specify the national organisation number or VAT number or a similar ID of the organisation. While we see the point in this, we feel it needs some more discussion. It can be added to the organisation object after its introduction. 6. changing the meaning of '-r' flag. The proposal suggested that, when a query returns an object with an "org:" attribute, the organisation objects mentioned in this attribute be appended to the query results. This behaviour can be changed by using '-r' flag in the query. This idea was based on the fact that the "org:" attribute is at least as relevant to the resource as the "admin-c:", "tech-c:" and "zone-c:" attributes as contact information. Thus, if we return person/role objects mentioned in the query results, we should also return the organisation objects. Basically, the proposed behaviour maintains the consistency. During the discussions it was proposed to change the default behaviour of whois server, and change the meaning of '-r' flag, so that - the server does not 'expand' "admin-c:", "tech-c:" and "zone-c:" attributes, nor "org:" attribute by default, - and it expands them only when '-r' flag is used. This can separately be discussed, however, I think this is out of the scope of organisation object. I hope I did not miss any other issues. If these points are OK, we will prepare another proposal for organisation object and publish it. Best regards, -- Engin Gunduz RIPE NCC Database Group

1 0

summary of organisation object discussions
by Engin Gunduz 08 Sep '03

08 Sep '03

Dear Colleagues, I'd like to summarise the discussions took place in the last two weeks in the working group mailing lists after the publication of the organisation object proposal ( http://www.ripe.net/ripe/mail-archives/db-wg/2003/msg00159.html ) and after the presentation in the DB-WG session last week in RIPE 46 Meeting ( http://www.ripe.net/ripe/meetings/ripe-46/presentations/ripe46-db-organisat… ). 1. "org:" attribute mandatory or optional in aut-num objects? The proposal mentioned above specified "org:" attribute in aut-num objects to be mandatory. However, often LIRs request an ASN for their customers and the RIPE NCC does not have enough information for them to fill in an organisation object completely. Rather than putting invalid information on the whois database, the RIPE NCC thinks it is better to make "org:" attribute optional in aut-nums. 2. "org:" attribute multi- or single-valued? The proposal specified "org:" attribute single-valued in all objects. During the discussions in the DB WG, commentors from the audience stated that in some cases a person object works for more than one companies, thus one might want to put multiple "org:" attributes in his/her person object. Similar reasoning could apply to other object types as well. However, the "org:" object specifies the holder of an Internet resource in inetnum, inet6num and aut-num objects, thus it must be single-valued in these objects. For the rest of the objects, we propose to modify the rule so it can be multi-valued. 3. addition of "mnt-ref:" attribute. The proposal suggested the use of "mnt-by:" attribute for both authorisation of modification of the organisation object and the authorisation of adding a reference to the organisation object from other objects. This has a shortcoming when the organisation object is protected by an entity other than the organisation itself (for example, an LIR, whose organisation object is protected by the RIPE NCC). We would propose to introduce mandatory "mnt-ref:" attribute to control the references to the organisation object, while "mnt-by:" protects the organisation object itself, as usual. 4. "org:" attribute in the organisation object itself. The proposal suggested the use of "org:" attribute in non-organisation objects. I think it could be used in organisation objects as well, to indicate business relations like being a dependent company to another one. 5. "business-id:" attribute. It was suggested to add an attribute to specify the national organisation number or VAT number or a similar ID of the organisation. While we see the point in this, we feel it needs some more discussion. It can be added to the organisation object after its introduction. 6. changing the meaning of '-r' flag. The proposal suggested that, when a query returns an object with an "org:" attribute, the organisation objects mentioned in this attribute be appended to the query results. This behaviour can be changed by using '-r' flag in the query. This idea was based on the fact that the "org:" attribute is at least as relevant to the resource as the "admin-c:", "tech-c:" and "zone-c:" attributes as contact information. Thus, if we return person/role objects mentioned in the query results, we should also return the organisation objects. Basically, the proposed behaviour maintains the consistency. During the discussions it was proposed to change the default behaviour of whois server, and change the meaning of '-r' flag, so that - the server does not 'expand' "admin-c:", "tech-c:" and "zone-c:" attributes, nor "org:" attribute by default, - and it expands them only when '-r' flag is used. This can separately be discussed, however, I think this is out of the scope of organisation object. I hope I did not miss any other issues. If these points are OK, we will prepare another proposal for organisation object and publish it. Best regards, -- Engin Gunduz RIPE NCC Database Group

1 0