- db-wg - mailman.ripe.net

new "via" attributes in aut-num Class
by Aris Lambrianidis 14 Aug '13

14 Aug '13

> Dear colleagues, > > We received the suggested patch to the RIPE Database software from Job Snijders containing the implementation of the features he describes below. The code he submitted meets the quality standards we have in the Database team and from a technical point of view we are willing to apply the patch. > > If the community supports the addition of the proposed attributes the RIPE NCC can go ahead and merge his implementation and make it available in the next release. > > Kind regards, > > Johan Åhlén > Assistant Manager Database > RIPE NCC > > On 14 Aug 2013, at 15:50, Job Snijders <job.snijders at atrato.com> wrote: > Hello Everybody, I think this is a nice, clean way to work towards ratifying some much needed route servers specific attributes, so this is a +1 from me for the patch to be merged. -- Best regards, Aris Lambrianidis AMS-IX NOC http://www.ams-ix.net/

1 0

new "via" attributes in aut-num Class
by Job Snijders 14 Aug '13

14 Aug '13

Dear working group, As some of you might know I've been working on extending the RPSL syntax to make life easier for certain scenario's, such as operating a Route Server. The purpose of the new attributes and the patch is to make life easier for Route Server operators whom have to deal with participants who have been assigned a 32 bit ASN. There currently is no way to relay information between non-adjacent ASNs, and 32 bit ASNs do not fit into standard BGP communities when coupled with an action code. Often Route server operators overload BGP communities to assess which participant should receive what prefixes, the "via" feature addresses exactly that concern. I have submitted the following draft to the IETF: http://tools.ietf.org/html/draft-snijders-rpsl-via And I have submitted a patch for the RIPE whois server to support the feature. https://github.com/job/whois The extension should be backward compatible with minimal impact on existing tools and processes, following Section 10.2 of RFC2622. RPSL parsers which do not understand the new attributes are expected to just ignore it. I am looking for feedback from this community, and especially the community's blessing to allow RIPE NCC to adopt the patch. Kind regards, Job

2 1

Draft proposal: Guidance Requested: Reassigning Referenced ASNs
by Ingrid Wijte 14 Aug '13

14 Aug '13

Dear colleagues, Thank you for your feedback on this issue. Based on the input we have received we propose taking the following action: - The RIPE NCC will send an email to the maintainer and contacts of the RIPE Database objects in which the returned AS numbers are referenced, asking for the reference to be removed. - The RIPE NCC will send three additional reminders, at intervals of three weeks from each other, for a total of four emails. - If after four emails the reference is still in place, the RIPE NCC will update the RIPE Database object itself, removing the reference. - If the reference is being recreated within one month from deletion (possibly due to automatic updates), the RIPE NCC will contact the organisation through means of communication other than email. The total number of RIPE Database objects that reference returned AS numbers is currently about 2000. All references are in the policy attributes of aut-num objects. No route(6) objects are involved. Thank you again and best regards, Ingrid Wijte RIPE NCC

2 1

New RIPE Database Release Available Soon
by Johan Åhlén 08 Aug '13

08 Aug '13

Dear colleagues, We are pleased to announce that a new release of the RIPE Database will be available soon. Details about the upcoming release, including dates, descriptions of the changes and other factors that might affect you as a user, can be found here: https://www.ripe.net/data-tools/db/release-notes/ripe-database-release-1.67… The first new feature is a diff functionality, which allows you to see, in detail, what changes were made between two versions of an object in the RIPE Database. Head over to RIPE Labs to read more about the diff feature: https://labs.ripe.net/Members/denis/diff-functionality-in-the-ripe-database We have also improved the way the Global Resource Service (GRS) data is imported. This allows us to present a more complete data set for all Internet resources throughout the world. This RIPE Labs article will bring you up to speed on the GRS improvements: https://labs.ripe.net/Members/denis/the-ripe-global-resource-service Lastly, we have also made several improvements to the REST API. As there are some changes, the new implementation will be deployed and running in parallel to the old version up until the next release. This provides you with time to verify your software and scripts against the new release. The implications of these changes are described in the release notes linked above. If you have any questions or comments, we look forward to hearing them! Kind regards, Johan Åhlén Assistant Manager Database RIPE NCC

2 2

abuse-c + org
by Gilles Massen 04 Jul '13

04 Jul '13

Dear WG, and RIPE NCC staff, While trying to add an abuse-c to our resources, I was quite surprised that you can only attach the abuse-c to an org object (while the policy suggests otherwise and implementation notes are not very clear on the limitation). So I'd like to ask why this restriction exists? What is wrong with adding an abuse-c directly to an inet[6]num or aut-num? Best regards, Gilles -- Fondation RESTENA - DNS-LU 6, rue Coudenhove-Kalergi L-1359 Luxembourg tel: (+352) 424409 fax: (+352) 422473

8 15

New Procedure for RIPE Database Software Releases
by Johan Åhlén 28 Jun '13

28 Jun '13

Dear colleagues, We have recently received valuable input on how we can improve our release procedures for the RIPE Database software. In order to further reinforce our processes, we have defined a new procedure together with the Database Working Group (WG) Chairs that will address the concerns that have been raised about the stability of the Database software by the community. The purpose of the new procedure is to give you notice of what the next release will contain and how it might affect you as a user. No matter what the impact of a new feature is, an announcement will always precede the deployment to production and you will be given sufficient time to test your software. The procedure is not too far from how we have been working up until now, but will be more strictly defined and will hopefully eliminate any surprises. We propose to proceed as follows: - We will move to a four to six week release cycle for the Database software - There will be no more than three new features per release - Each new release will be announced to the WG mailing list at least two weeks before we deploy to production - In the event that we are required to deviate from what was stated in the announcement, we will announce this to the WG mailing list - The new release will be available in Test Database at least two weeks before we deploy to production - We will announce to the WG mailing list when we have successfully completed deployments to Test Database - We will announce to the WG mailing list when we have successfully completed deployments to Production Database - In the special case of required emergency fixes, we will announce this to the WG mailing list as well as the website as soon as possible - We will make a follow-up announcement to the WG mailing list and the website as soon as a fixed version has been deployed - These two emergency announcements can be combined for security reasons, or if the problem is solved in less than 30 minutes For the initial release announcements, the following template will be used: Release number Expected date of deployment to Test Database Expected date of deployment to Production Database New features Impact on existing functionality Operational improvements Impact on existing functionality Fixed bugs Impact on existing functionality Expected upcoming features for next release We hope that you will find this procedure beneficial and with your consent we would like to move ahead with it effective immediately. We also intend to publish this procedure to the website together with a roadmap of what we expect that the upcoming releases will contain. Kind regards, Johan Åhlén Assistant Manager Database RIPE NCC

1 0

DB Software - release and deployment improvements
by Wilfried Woeber 28 Jun '13

28 Jun '13

Dear DB-WG, NCC-Services and Routing WG Community! Over the recent month or so, a deployment of a new version of the DB-SW, to the production environment, turned out to (potentially) have serious impact on automated router configurations in (an) operator's network(s). The issue was reported to the NCC's DB-WG Group and the specific problem has already been fixed (actually on the same day as reported). Here's the link to the service announcement page: http://www.ripe.net/lir-services/service-announcements/ripe-database-query-… However, as the incident seemes serious enough, a small group of interested people (Rüdiger, Nigel Titley, folks from the NCC and myself) started to dig a bit deeper, into what was (and is) perceived to be aspects that need review, more transparency and maybe improvement. We plan to look at some aspects, amongste others, like: - better communication regarding upcoming new SW releases - better transparency when and where (TEST DB, Production DB,...) new software will be deployed - version control and documentation But the most important thing is an improved and more transparent process to deploy new software versions *as soon as possible*! Then we can take it from there... While this message is also sent to the NCC Services and Routing WG, fyi, I'd like to suggest to have follow-up comments and discussion on the DB-WG List! Looking forward to fruitful discussions, best regards, Wilfried

1 0

Re: [db-wg] [address-policy-wg] Cosmetic Surgery Project: Extended Review Period until 9 July on revised RIPE Policy document for IPv6 database objects
by Emilio Madaio 26 Jun '13

26 Jun '13

Dear Peter, thank you very much for your input and remark. I apologize if I forgot to include the IPv6 and Database WG mailing lists as decided during RIPE 66 on request of the community. I slavishly followed the procedure defined by the Cosmetic Surgery Project that was meant for the Address Policy WG mailing list only. The original announcement for the extended Review Phase is available at: http://www.ripe.net/ripe/mail/archives/address-policy-wg/2013-June/007931.h… For further feedback the draft of the policy document is online and ready for community review at: https://www.ripe.net/ripe/readability/improving-the-readability-of-ripe-doc… Please send your feedback on this draft document before 9 July 2013. Best Regards Emilio Madaio Policy Development Officer RIPE NCC On 6/25/13 5:51 PM, Peter Koch wrote: > On Tue, Jun 25, 2013 at 03:29:45PM +0200, Emilio Madaio wrote: > >> https://www.ripe.net/ripe/readability/improving-the-readability-of-ripe-doc… > > May I express some doubts about the 'cosmetic surgery project'? > The project was introduced as early as October 2009 and RIPE 513 was published > in February 2011. What appears to happen is a very late post publication copy > editing. In this particular case, the policy itself is a change to the > database attributes and - other than, say, address allocation/assignment > policies - not very likely read its own. Any structural changes > to the document (where's the template, by the way?) and changes to style > (passive voice here and there) or readability might better be invested > in the general database documentation. > > The new draft eliminates the data protection aspect from the > introduction/motivation section, which I consider a loss. The abstract > wins, though. The draft continues to use 2000::/46 for the example > where some chunk of 2001:DB8::/32 as per RFC 3849 might be a better choice. > > -Peter >

1 0

Guidance Requested: Reassigning Referenced ASNs
by Andrea Cima 25 Jun '13

25 Jun '13

[Apologies for duplicates] Dear colleagues, At the Address Policy WG session during RIPE 66, the RIPE NCC requested guidance from the RIPE community on reassigning referenced 16-bit ASNs. Together with the Address Policy WG Chairs, we have decided to seek additional input from the WG mailing list. Background: the RIPE NCC has around 1,400 returned 16-bit ASNs that are currently referenced in route objects, and "import" and "export" lines in aut-num objects. Problem: there is a need for 16-bit ASNs by operators. However, because these ASNs are referenced, they are not being reassigned. At RIPE 66, the RIPE NCC suggested the following potential solutions: A. Not reassign referenced AS numbers B. Reassign AS numbers even if referenced C. Clean up the RIPE Database from references and then reassign the AS numbers C.1. "Silent" update; or C.2. Informing the object holder of the update Please find the RIPE Meeting slides at the following url: https://ripe66.ripe.net/presentations/176-Address_Policy_WG_RIPE_66.pdf This email has been sent to both the Address Policy WG and RIPE Database WG mailing lists, due to the impact the proposals would have on the RIPE Database. However, we would ask to keep the discussion on the Address Policy WG mailing list to facilitate participation and the channeling of feedback. Thank you in advance for your input and best regards, Andrea Cima RIPE NCC

2 1

Object Tags in the RIPE Database
by Johan Åhlén 19 Jun '13

19 Jun '13

Dear colleagues, The recent RIPE Database software rewrite has enabled us to offer new and valuable functionality to the community with minimal effort. One of these features is the Object Tags functionality, initially introduced by Kaveh Ranjbar at RIPE 66 and since then communicated to the community and the Database Working Group Chairs. We are now happy to announce that the Object Tags feature is fully deployed. Object Tags is publicly visible metadata related to an object in the RIPE Database. The tags do not appear by default and users can enable them by supplying the --show-tag-info option. The tags will appear in the comment section and it does not change the default behaviour of your queries. The main goal of the tags is to improve the data quality in the RIPE Database by, for example, better notifying users of potential issues with an object. As a secondary benefit, query results can now be filtered based on tags, making it easier for users to find exactly what they need. The RIPE NCC will maintain a list of the available tags on a webpage referenced in the Help section and in related documentation. Currently, there are two types of tags: - Objects due to be deleted by the automated clean-up are tagged: Unreferenced # will be deleted in n days - All RIPE and GRS resource data is tagged with either: {SOURCE}-REGISTRY-RESOURCE or {SOURCE}-USER-RESOURCE For more information and examples of how the new feature can be used, I'd like to refer you to Kaveh's RIPE Labs article at: https://labs.ripe.net/Members/kranjbar/ripe-database-object-tags We hope that you can come up with creative ways of using this functionality. We are eager to hear how you use this feature and what we can do to make it better. I would also like to take this opportunity to introduce myself as not everyone here knows who I am. My name is Johan Åhlén and I'm the new Assistant Database Team Manager at the RIPE NCC. I'm taking over Kaveh's role of managing the Database Team as Kaveh has accepted the new position of Head of Science Division. Before managing the Database Team, I worked as a Senior Software Engineer in the Business Applications Department where I was involved in the development of both internal and external RIPE NCC applications. I look forward to working with you all. Kind regards, Johan Åhlén Assistant Database Team Manager RIPE NCC

1 0