- db-wg - mailman.ripe.net

Fw: FAILED:
by benhadjira 06 Oct '03

06 Oct '03

Dear sir We have tried several times to add inetnum to the data-base, but unfortunetely this is the error message we received every time . Could you please give us some help & advices on how to do it ? Thanking you in advance Mourad Techni-communication -------Original Message------- From: ripe-dbm(a)ripe.net Date: samedi 4 octobre 2003 10:34:28 To: benhadjira(a)techni-communication.com Subject: FAILED: > From: benhadjira(a)techni-communication.com > Subject: > Date: Sat, 04 Oct 2003 05:34:19 -0400 > Reply-To: benhadjira(a)techni-communication.com > Message-ID: <04df532f66b14c6c874ce4e412db3b3d(a)techni-communication.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: 1 Modify: 0 Delete: 0 Syntax Errors: 0 DETAILED EXPLANATION: ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ The following object(s) were found to have ERRORS: --- Create FAILED: [inetnum] 217.29.219.0 - 217.29.219.255 ***Error: Authorisation failed ***Info: Syntax check passed inetnum: 217.29.219.0 - 217.29.219.255 netname: Techni-communication descr: Techni-Communication - customer SEI.NET country: DZ admin-c: MA1973-ripe tech-c: BM1974-RIPE status: LIR-PARTITIONED PA mnt-by: BM-MNT mnt-lower: BM-MNT mnt-routes: TACHYON-EU-MNT changed: benhadjira(a)techni-communication.com 20031004 source: RIPE ***Info: Authorisation for parent [inetnum] 217.29.208.0 - 217.29.223.255 using mnt-lower: not authenticated by: BM-MNT ***Info: Authorisation for [inetnum] 217.29.219.0 - 217.29.219.255 using mnt-by: not authenticated by: BM-MNT ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ The following paragraph(s) do not look like objects and were NOT PROCESSED: -------------------------------------------------------------------- mail2web.com - Check your email from the web at http://mail2web.com. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For assistance or clarification please contact: RIPE Database Administration <ripe-dbm(a)ripe.net> .

2 1

Reverse DNS Restructuring Project
by Olaf M. Kolkman 01 Oct '03

01 Oct '03

Dear Colleagues, I have just posted a proposal about the restructuring of the reverse DNS setup to the ncc-services-wg mailing list. I'd like to point out that this proposal has relevance to both the dns-wg and db-wg. You can find the proposal on http://www.ripe.net/reverse/proposal.html Please provide your feedback on the ncc-services-wg mailing list. With excuses for the cross-post, -- Olaf ---------------------------------| Olaf M. Kolkman ---------------------------------| RIPE NCC

1 0

Looking for good way to find blocks allocated directly by ripe
by william＠elan.net 27 Sep '03

27 Sep '03

Hi, I'd like to find what is the best way to determine if particular inetnum was allocated directly by RIPE, this would include all blocks allocated to LIRs, all blocks allocated to large organizations and blocks allocated as small PI blocks. Currently I have come with the following way to do it - first find all inetnum objects listed as "ALLOCATED PI" with network name starting as "EU-ZZ", this find all PI space; then discard all "ASSIGNED PA" blocks except those within range of previously found "ALLOCATED PI" blocks and also discard "ALLOCATED PI" and "LIR-PARTITIONED PA" blocks. >From reminder the netblocks totally within another inetnum are discarded as well as "superblocks" listed under RIPE name. I'd like to find if you think doing it this way would leave any blocks out or include too many larger block then direct RIPE allocations. Doing it the way I described above, already produced some data available at http://www.completewhois.com/bogons/data/allocated-cidr-ripe_main.txt and non-allocated space is at http://www.completewhois.com/bogons/data/bogons-cidr-ripe_main.txt I'd like to verify validity of this data and the process I used, especially considering that I see number of errors, for example there are several blocks > /24 listed as allocated (but I did not think RIPE did any direct allocations or assignemsns of size > /24), for example: 193.58.0.64/26 193.164.232.192/26 193.201.147.0/26 193.201.148.64/26 193.243.183.0/26 193.254.0.64/26 194.1.127.128/26 194.9.166.0/26 194.149.71.128/26 194.153.156.0/26 194.180.159.128/26 194.180.226.192/26 194.246.39.0/26 194.246.39.128/26 195.35.104.64/26 -- William Leibzon Elan Networks william(a)elan.net

3 3

Organisation Object Proposal, version 3
by DB-News 22 Sep '03

22 Sep '03

Dear Colleagues, [apologies for duplicate messages] After the discussions we had in the RIPE 46 Meeting and in the WG mailing lists, we have incorporated agreed modifications into organisation object proposal, which you can find below. This is the Version 3 of the proposal. You can find the first version at http://www.ripe.net/ripe/mail-archives/db-wg/2003/msg00013.html the second version at http://www.ripe.net/ripe/mail-archives/db-wg/2003/msg00159.html and a summary of discussions on the second version at http://www.ripe.net/ripe/mail-archives/db-wg/2003/msg00213.html Thank you very much for your input on the issue. 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 objects using an "org:" attribute. All objects associated with a particular organisation ID can be retrieved by performing an "inverse query" for this ID 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:", "mnt-ref:", "organisation:", "org-name:", "org-type:" and "org:" have their usual meanings. organisation: [mandatory] [single] [primary/look-up key] org-name: [mandatory] [single] [look-up key] org-type: [mandatory] [single] [ ] descr: [optional] [multiple] [ ] remarks: [optional] [multiple] [ ] address: [mandatory] [multiple] [ ] country: [mandatory] [single] [ ] phone: [optional] [multiple] [ ] fax-no: [optional] [multiple] [ ] e-mail: [mandatory] [multiple] [look-up key] org: [optional] [multiple] [inverse key] admin-c: [mandatory] [multiple] [inverse key] tech-c: [mandatory] [multiple] [inverse key] ref-nfy: [optional] [multiple] [inverse key] mnt-ref: [mandatory] [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. "mnt-ref:" Specifies the maintainer objects that are entitled to add references to the organisation object from other objects. "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 object type. It points to an existing organisation object representing the entity that holds the resource, in the cases where the whois DB object represents an Internet resource. In other objects, it can be used to specify the business relations. The value of this attribute is the ID of the organisation object. It is mandatory in the 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. "org:" attribute is single-valued in the inetnum, inet6num and aut-num objects, and it is multi-valued all other objects. The "org:" attribute is used to specify the holder of a resource in inetnum, inet6num and aut-num objects, thus it must be single-valued in them. In other objects, it specifies business relations (like in a person object, where it can be used to specify whom the person works for) it can be multiple (in the person object example, a person might work for several companies). 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 in the "mnt-ref:" attributes of the organisation object - from one of the maintainers in the "mnt-by:" attributes 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. 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-ref: EXAMPLE-MNT 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

1 0

Re: AW: [local-ir@ripe.net]Proposed change of "mnt-lower:" behaviour: procedure and timelines
by Katie Petrusha 18 Sep '03

18 Sep '03

On Thu, Sep 18, 2003 at 01:36:16PM +0200, Bornhorst, Ludger wrote: Dear Ludger, > > To solve this problem, allocation objects have to be modified to include a > > "mnt-lower" attribute. We have applied certain heuristics to determine the most > > suitable maintainer for an allocation. If there is no maintainer, a new > > maintainer will be generated. > > in the case of PA allocations I agree with you, to include a "mnt-lower" > attribute to all PA allocation objects. > > But what about PI Allocations? In order to create an appropriate route object > for PI addresses, it is necessary to update (not create!) the ASSIGNED PI- > inetnum with a mnt-routes attribute etc. to get through the authorization > for the route object creation. The proposed change in the behaviour won't affect the updates of the existing objects. It affects only the _creation_ of more-specific inetnums (inet6nums, domains). > Is it guaranteed, that we can perform ASSIGNED PI-inetnum updates in the future? Yes, as long as the update is authorised by "mnt-by:" of the object itself. Hopefully this answers your question. > > regards > Lu > > Ludger Bornhorst > ______________________________________________________________________ > > Deutsche Telekom AG > T-Com Headquarters > Network Information Center > Ammerlander Heerstr. 138, D-26129 Oldenburg Hotline > +49 441 234 4581 (phone) +49 800 3301180 > +49 441 234 4589 (fax) +49 800 3301179 > ludger.bornhorst(a)telekom.de (mail) dk.call-center(a)telekom.de > -- Katie Petrusha Database Group RIPE NCC

1 0

Proposed change of "mnt-lower:" behaviour: procedure and timelines
by Katie Petrusha 18 Sep '03

18 Sep '03

[Apologies for duplicate messages] Dear colleagues, In March 2003, the RIPE NCC circulated the proposal to change the behaviour of the "mnt-lower:" attribute in inetnum, inet6num and domain objects. Currently, if the "mnt-lower" attribute of a parent object is absent, anybody can create more-specific objects. The goal is to change this behaviour to a more secure Routing Policy System Security (RPSS) style scheme. In RPSS, objects use "mnt-lower:" to specify a maintainer which has the ability to authorise the creation of more-specific objects. If a "mnt-lower:" attribute is not present, then the "mnt-by:" of the less-specific object is used. A similar scheme is already used for route object creation ("mnt-routes:" attribute). As inetnum objects representing allocations are maintained by the RIPE NCC, deployment of this scheme may not allow LIRs to create assignments from their allocation if the LIR does not have a "mnt-lower:" attribute pointing to the LIR's maintainer. To solve this problem, allocation objects have to be modified to include a "mnt-lower" attribute. We have applied certain heuristics to determine the most suitable maintainer for an allocation. If there is no maintainer, a new maintainer will be generated. Details about project assumptions and heuristics were presented at RIPE 46: http://www.ripe.net/ripe/meetings/ripe-46/presentations/ripe46-db-allocatio… The initial proposal was sent in March 2003: http://www.ripe.net/ripe/mail-archives/db-wg/2003/msg00033.html The RIPE NCC suggests the following implementation plan: Step 1. Distribute this plan for approval. Time: 2 weeks Step 2. Notify contacts whose allocations need to be modified Wait for feedback and people making updates themselves Time: 1 month Step 3. RIPE NCC will check all affected allocations again and update the ones still requiring modification, generating necessary maintainers and notifying contacts. Time: 1 day Step 4. Passwords for generated maintainers made available through the LIR Portal. Wrong modifications can be corrected through the LIR Portal as well. Time: 1 day Step 5. Deploy the change in the RIPE Whois Database Software to support new scheme. Time: to be announced Any suggestions or comments on this plan are welcome. If your allocation doesn't have a "mnt-lower:" attribute, we encourage you to update it as soon as possible through the LIR portal. If you don't have a "mnt-lower:" after this proposal has been implemented, you won't be able to create any new assignments for your allocations. We believe that this change will be beneficial for all RIPE Whois Database users and we hope the migration will go as smoothly as possible. Thanks for your co-operation. Katie Petrusha Database Group RIPE NCC

1 0

Re: SUCCESS: Bulk deletion of /24 detail
by Niall O'Reilly 15 Sep '03

15 Sep '03

[Input to DB-WG] Flagging this result as 'SUCCESS' is unhelpful. I regard this as a bug, and a silly one too. [Question to ripe-dbm] So how am I supposed to delete redundant objects and keep the DB clean ? Best Niall O'Reilly On Friday, Sep 12, 2003, at 08:26 Europe/Dublin, ripe-dbm(a)ripe.net wrote: > > >> From: "Niall O'Reilly" <Niall.oReilly(a)ucd.ie> >> Subject: Bulk deletion of /24 detail >> Date: Fri, 12 Sep 2003 08:26:41 +0100 >> Reply-To: Niall.oReilly(a)ucd.ie >> Message-ID: <7484C0E0-E4F2-11D7-A9BD-000393DA759C(a)ucd.ie> > > > SUMMARY OF UPDATE: > > Number of objects found: 0 > Number of objects processed successfully: 0 > Create: 0 > Modify: 0 > Delete: 0 > No Operation: 0 > Number of objects processed with errors: 0 > Create: 0 > Modify: 0 > Delete: 0 > Syntax Errors: 0 > > > DETAILED EXPLANATION: > > > ***Warning: Invalid keyword(s) found: Bulk, deletion, of, /24, detail > ***Warning: All keywords were ignored > > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > The following paragraph(s) do not look like objects > and were NOT PROCESSED: > > inetnum: 193.1.128.0 - 193.1.128.255 > netname: UCD-193-1-128-19-128 > delete: aggregated into UCD-193-1-128-19 > > inetnum: 193.1.129.0 - 193.1.129.255 > netname: UCD-193-1-128-19-129 > delete: aggregated into UCD-193-1-128-19 > > inetnum: 193.1.130.0 - 193.1.130.255 > netname: UCD-193-1-128-19-130 > delete: aggregated into UCD-193-1-128-19 > > inetnum: 193.1.131.0 - 193.1.131.255 > netname: UCD-193-1-128-19-131 > delete: aggregated into UCD-193-1-128-19 > > inetnum: 193.1.133.0 - 193.1.133.255 > netname: UCD-193-1-128-19-133 > delete: aggregated into UCD-193-1-128-19 > > inetnum: 193.1.137.0 - 193.1.137.255 > netname: UCD-OLDBAR-LHS > delete: aggregated into UCD-193-1-128-19 > > inetnum: 193.1.138.0 - 193.1.138.255 > netname: UCD-OLDBAR-RHS > delete: aggregated into UCD-193-1-128-19 > > inetnum: 193.1.139.0 - 193.1.139.255 > netname: UCD-MICROSTORE > delete: aggregated into UCD-193-1-128-19 > > > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > > > For assistance or clarification please contact: > RIPE Database Administration <ripe-dbm(a)ripe.net> > >

4 6

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