- db-wg - mailman.ripe.net

Re: [db-wg] WG Chair Selection - Call for Discussion
by Andy Davidson 06 Sep '18

06 Sep '18

Hi, William Thank you kindly for working on this process. Please could you confirm how many seats are open for selection this time? Are we selecting for one seat or two from this list of people? Andy

2 1

Status Update on NWI-5 Deployment: Completed
by Nathalie Trenaman 05 Sep '18

05 Sep '18

Dear colleagues, On 4 September, 21:55 UTC+2, we completed the following changes needed for NWI-5: Out of region ROUTE(6)/AUT-NUM objects. - Removal of mnt-routes from AUT-NUM objects - Removal of the RIPE-NCC-RPSL-MNT - Update the source attribute for out-of-region ROUTE(6) objects to RIPE-NONAUTH On 5 September, 04:32 UTC+2, we also completed the new database dump and split files at: ftp://ftp.ripe.net/ripe/dbase/ <ftp://ftp.ripe.net/ripe/dbase/> We have updated the release notes accordingly: https://www.ripe.net/manage-ips-and-asns/db/release-notes/ripe-database-rel… <https://www.ripe.net/manage-ips-and-asns/db/release-notes/ripe-database-rel…> Of course, we will keep monitoring the impact on operators. We would like to thank everyone who contributed with valuable input during the test and deployment phases. Kind regards, Nathalie Trenaman Product Manager RIPE NCC

2 1

Implementation plan and time line for NWI5 - Out-of-region ROUTE(6) objects and removal of ASN authorisation for ROUTE(6) object creation.
by Nathalie Trenaman 05 Sep '18

05 Sep '18

Dear colleagues, On Tuesday, 4 September, we will make changes to the way out-of-region objects are handled in the RIPE Database, as directed by the RIPE Database Working Group. We will implement the following changes in this order between 09:00 and 15:00 UTC+2. 1. The RIPE Routing Registry will no longer support the creation of out-of-region ROUTE(6) and AUT-NUM objects 2. The creation of ROUTE(6) objects will no longer need authorisation from the ASN holder 3. Existing out-of-region objects will have their "source:" attribute changed to "RIPE-NONAUTH” For users of NRTM, these changes will be reflected on the FTP website after 00:00 UTC+2 on Wednesday, 5 September. You can find our implementation plan at: https://www.ripe.net/manage-ips-and-asns/db/impact-analysis-for-nwi-5-imple… <https://www.ripe.net/manage-ips-and-asns/db/impact-analysis-for-nwi-5-imple…> We also published an article on RIPE Labs with some background and additional context: https://labs.ripe.net/Members/denis/out-of-region-route-6-and-aut-num-objec… <https://labs.ripe.net/Members/denis/out-of-region-route-6-and-aut-num-objec…> We will inform the Database Working Group once all changes have been implemented. Kind regards Nathalie Trenaman Product Manager RIPE NCC

2 2

Status Update on NWI-5 Deployment
by Nathalie Trenaman 04 Sep '18

04 Sep '18

Dear colleagues, We are currently deploying the changes for NWI-5 in the RIPE Database. At the moment, we have successfully rolled out: - the removal of mnt-routes from AUT-NUM objects - removal of the RIPE-NCC-RPSL-MNT The rollout of updating the “source:" attribute to RIPE-NONAUTH on out-of-region ROUTE, ROUTE6 and AUT-NUM objects is taking slightly longer than expected. This means that the source will gradually be updated on out-of-region objects for the rest of today as the script runs. Kind regards, Nathalie Trenaman Product Manager RIPE NCC

1 0

Increasing the Limit of IP Addresses Used for the NRTM Service
by Theodoros Polychniatis 30 Aug '18

30 Aug '18

Dear colleagues, We would like to increase the limit of IP addresses used with the NRTM service. The limit of two addresses per member was set eight years ago and we might be able to handle up to three times that amount now. We are in the process of determining the new limit. If you are interested in helping us test the capacity of our NRTM servers, please open a ticket with us to add more IP addresses to the ones you currently use: https://www.ripe.net/contact-form?topic=ripe-dbm&show_form=true Kind regards, Theodoros Polychniatis Manager SW Engineering RIPE NCC

1 0

improve --list-sources query to display latest serial?
by Job Snijders 30 Aug '18

30 Aug '18

Hi all, It appears there currently is no easy way to retrieve the current serial from the RIPE DB. We're implementing some software to monitor whether our local mirror is wildly out of sync with the RIPE DB or not. I kind of expected th serial to be listed in the --list-sources command, but that field is set to 0. hanna:~ job$ whois -h whois.ripe.net -- --list-sources | grep RIPE % This is the RIPE Database query service. % The RIPE Database is subject to Terms and Conditions. RIPE:3:N:0-0 RIPE-GRS:3:N:0-0 % This query was served by the RIPE Database Query Service version 1.91.2 (HEREFORD) In comparison, IRRd will display the current serial when the !j query is issued: hanna:~ job$ whois -h rr.ntt.net -- '!jNTTCOM' A25 NTTCOM:N:0-978492:978492 C hanna:~ job$ I came up with a really ugly oneliner: $ whois -h 193.0.6.135 -p 4444 -- \ "-g RIPE:3:$(curl -s https://ftp.ripe.net/ripe/dbase/RIPE.CURRENTSERIAL)-LAST" | \ grep -v : | sed '/^$/d' | tail -n 2 | head -1 | awk '{ print $NF }' 42486238 Depending on the time of day that oneliner may be quite quick, or take several minutes to complete... obviously not ideal. Can --list-sources perhaps be improved to list the current serial? Kind regards, Job

2 2

improve --list-sources query to retain history
by Erik Bais 30 Aug '18

30 Aug '18

If the WG is going to look at improving the --list-sources code base, I have something that I would love to see on the wish list to be implemented.. Maintaining history (upto the point where a prefix is leaving the RIPE DB, due to an inter-rir transfer. ) is helpful to see who the previous holder was or where the prefix came from. This is particular the case with Legacy prefixes that are being split up.. as it is current practise to delete the parent prefix after a prefix split and transfer. Once the parent is deleted, you can't list or request the history of the parent prefix anymore. If the historic --list-versions and --show-version could be preserved somehow (on the original prefix size.. instead of them being deleted..), that would be very helpful. Regards, Erik Bais

2 1

NWI-5 changes deployed to RC (Release Candidate) environment
by Nathalie Trenaman 30 Aug '18

30 Aug '18

Dear Colleagues, Following the implementation plan for NWI-5 (Out-of-region ROUTE(6) objects and removal of ASN authorisation for ROUTE(6) object creation) we have now deployed the NWI-5 changes to the RC (Release Candidate) environment. The RC environment allows you to test the impact of the changes before we will go to production on 4 September 2018. Changes include: - removal of mnt-routes from aut-num objects - updating the source: attribute to RIPE-NONAUTH on out-of-region route,route6 and aut-num objects - removal of the RIPE-NCC-RPSL-MNT In this RC release, the FTP split files, the database dump and the NRTM stream will contain objects with the sources RIPE and RIPE-NONAUTH. In the next release (in the coming weeks), there will be two NRTM streams, two database dumps and also two sets of split files, one per source. We will separately announce this release to the Database Working Group. We encourage you to test the impact of the NWI-5 changes on your scripts and tools and report any issues you may find to ripe-dbm(a)ripe.net <mailto:ripe-dbm@ripe.net> Information about Release Candidate can be found at: https://www.ripe.net/manage-ips-and-asns/db/release-notes/ripe-database-rel… <https://www.ripe.net/manage-ips-and-asns/db/release-notes/ripe-database-rel…> The full NWI-5 implementation plan can be found at: https://www.ripe.net/manage-ips-and-asns/db/impact-analysis-for-nwi-5-imple… <https://www.ripe.net/manage-ips-and-asns/db/impact-analysis-for-nwi-5-imple…> Kind regards, Nathalie Trenaman Product Manager

2 2

Incorrect RDAP response
by Brian Rak 25 Aug '18

25 Aug '18

https://rdap.db.ripe.net/ip/24.235.22.0/23 This will just infinitely redirect between ARIN and RIPE: $ wget https://rdap.arin.net/registry/ip/24.235.22.0/23 --2018-08-24 14:00:05-- https://rdap.arin.net/registry/ip/24.235.22.0/23 Resolving rdap.arin.net... 2001:500:a9::160, 2001:500:31::160, 2001:500:13::160, ... Connecting to rdap.arin.net|2001:500:a9::160|:443... connected. HTTP request sent, awaiting response... 303 See Other Location: https://rdap.db.ripe.net/ip/24.235.22.0/23 [following] --2018-08-24 14:00:06-- https://rdap.db.ripe.net/ip/24.235.22.0/23 Resolving rdap.db.ripe.net... 2001:67c:2e8:22::c100:68e, 193.0.6.142 Connecting to rdap.db.ripe.net|2001:67c:2e8:22::c100:68e|:443... connected. HTTP request sent, awaiting response... 301 Moved Permanently Location: https://rdap.arin.net/registry/ip/24.235.22.0/23 [following] --2018-08-24 14:00:06-- https://rdap.arin.net/registry/ip/24.235.22.0/23 Connecting to rdap.arin.net|2001:500:a9::160|:443... connected. HTTP request sent, awaiting response... 303 See Other Location: https://rdap.db.ripe.net/ip/24.235.22.0/23 [following] --2018-08-24 14:00:07-- https://rdap.db.ripe.net/ip/24.235.22.0/23 Connecting to rdap.db.ripe.net|2001:67c:2e8:22::c100:68e|:443... connected. HTTP request sent, awaiting response... 301 Moved Permanently Location: https://rdap.arin.net/registry/ip/24.235.22.0/23 [following] --2018-08-24 14:00:07-- https://rdap.arin.net/registry/ip/24.235.22.0/23 Connecting to rdap.arin.net|2001:500:a9::160|:443... connected. HTTP request sent, awaiting response... 303 See Other It seems to be a RIPE issue though, because RIPE has valid data when I query via the web UI: https://apps.db.ripe.net/db-web-ui/#/query?searchtext=24.235.22.0%2F23

2 1

The Ongoing Summer of Hijacks: MNT-SERVERSGET / dnsget.top
by Ronald F. Guilmette 23 Aug '18

23 Aug '18

The entire set of objects in the RIPE WHOIS data base that are currently registered with mnt-by: MNT-SERVERSGET is listed here: https://pastebin.com/raw/GiYWxHMh Among this set of objects there are 235 separate route objects. Evidence indicates persuasively that some sizable fraction of these RIPE- registered route objects are fradulent and are simply there to provide cover for multiple IPv4 address block hijacks. The presence of these objects in the data base permits the following set of ASNs to claim that they are acting "legitimately" even as they route these hijacked blocks: AS9009 M247 Ltd (UK) AS43350 NFOrce Internet Sevices (Netherlands) AS57129 Optibit, LLC (Russia) AS197328 Istanbuldc Veri Merkezi Ltd. Sti (Turkey) AS202287 Men Danil Valentinovich (Russia) AS204895 Santa Plus, LLC (Russia) The total amount of IPv4 space encompassed within the set of route objects registered with mnt-by: MNT-SERVERSGET at the present time amounts to eight hundred and fifty nine (859) /24 blocks. Of these, only three hundred and five (305) actually have correctly functioning and properly delegated reverse DNS at the present time, and even among those, only two hundred and two (202) have functioning reverse DNS delegations to the prefered name servers of MNT-SERVERSGET, which is to say the name servers ns5.dnsget.top and ns6.dnsget.top. The bottom line is that it appears that, at the present time, something less than 1/4 of all of the IPv4 address space currently registered in the RIPE data base (via route objects) by and to MNT-SERVERSGET is space for which a plausible case could be made that the blocks in question are actually legitimately assigned to and/or under the legitimate control of whoever or whatever is MNT-SERVERSGET. The other 3/4ths of the IPv4 space in question has provenance which is, at best, dubious. Due to its use of little country-of-registration flags for each IP address block, the web site bgp.he.net provides the most visually obvious indications of at least two of the specific block hijacks in this case, specifically the hijacks of 27.103.192.0/19 and 36.0.192.0/19 by AS57129: https://bgp.he.net/AS57129#_prefixes Based upon the foregoing, I hereby respectfully request RIPE NCC to undertake an immediate and conmprehensive review of all objects in the data base that are currently registered with mnt-by: MNT-SERVERSGET. Additionally, I also respectfully request RIPE NCC to publish the results of this review to the mailing lists of the Database Working Group and the Anti-Abuse Working Group. The charters of both of these Working Groups are directly relevant to this issue, and there exists neither need nor reason to simply sweep this issue quietly under the carpet, as has been done in previous and similar cases. Cases such as this, and the two others of similar magnitude that I have publicly disclosed just this summer, affect the operation of, the stability of, and the continued enjoyment of the entire planetary Internet and its billions of users. Despite its status as a strictly private corporation, RIPE has a public responsibility, not only to handle such incidents responsibly and competently, but to show the world that it is ready, willing, and able to do so. The responses of RIPE to prior instances of this exact type of bad behavior have been largely or entirely cloaked in secrecy, presumably to protect the guilty. This longstanding and antiquated tradition of omerta within the RIPE community is unambiguously counterproductive to the goal of a well-managed and properly fuctioning Internet. Just as importantly, it naturally gives rise to unavoidable questions about the actual competence of, and capabilities of RIPE NCC staff as they attempt to deal with such incidents. I personally feel that RIPE NCC staff are doing the best they can when responding to incidents such as this, but the tradition of playing "hide the ball" with respect to their actions in such cases reflects badly on them, and badly on RIPE generally. It certainly raises doubts about RIPE's claim to authority over even its own data base. Lastly, I respectfully request both the RIPE Executive Board and the RIPE membership to make plain and explicit their respective intentions with regards to the various bad actors that have been caught, and that may in future be caught red handed engaging in the deliberate and premeditated corruption of the RIPE data base. To date, the policies an actions that RIPE applies, or which RIPE may apply to such bad actors have been shrouded in apparently deliberate secrecy and mystery. To put this request in more concrete terms, I would like to know if RIPE and the Executive Board actually and deliberately intend to take no action whatsoever with respect to the ongoing RIPE memberships of clearly identified bad actors, specifically, in this case, whatever persons or entities are currently hiding behind the RIPE WHOIS handle MNT-SERVERSGET (aka AS202287 and AS202275), which would appear to be this specific entity / RIPE member: https://www.ripe.net/membership/indices/data/ru.danil.html Is it the express intention of both the Board and the RIPE membership to simply deprive this bad actor of just those fradulent data base entries that have effectively legitimized his/her/its fradulent routing announcements? Is it the express intention of both the Board and the RIPE membership to simply make this bad actor give back what he has stolen, and to otherwise take no action, just as the response has been in the two other and similar cases that have also arisen in the RIPE region and that I have also publicly reported on this summer? https://mailman.nanog.org/pipermail/nanog/2018-June/096034.html https://mailman.nanog.org/pipermail/nanog/2018-July/096437.html My question is prompted by the following simple facts. In the two prior cases cited above, and also in the one I am presenting here today, the bad actors involved were seen to have not only hijacked large swaths of IP address space that clearly didn't belong to them, but also, as in the case I am presenting today, these same bad actors additionally acted to corrupt the RIPE data base with premeditated and deliberately fradulent entries. Nonetheless, and regardless of these attacks on the reliability and trustworthyness of the RIPE data base, to date it appears that no action whatsoever has been taken which might affect the ongoing RIPE memberships of the relevant parties and bad actors involved, and they are all still members in good standing of RIPE at the present time: https://www.ripe.net/membership/indices/data/pt.bitcanal-pt.html https://www.ripe.net/membership/indices/data/ua.d2investukraine.html https://www.ripe.net/membership/indices/data/bz.universal.html This is, to say the least, puzzling. I would like to know when, where, and how the Executive Board and/or the RIPE membership reached the altogether dubious conclusion that the best possible way to deal with bad actors such as these is to make them give back -just- the stuff the stole, and then to otherwise impose no penality of any kind, thus allowing them to live on, so that they may hijack another day. Whether this policy is a result of either deliberation or default, it *does* appear to be the policy, based on the evidence. Without intending to be rude, I feel that I must ask the obvious question: In what Universe does this otherwise admirable and generous Christian policy of "turning the other cheek" with respect to such verified bad actors have any effect other than encouraging the next hijacker, and then next one, and the one after that? Has either the Board or the membership even ever seriously debated what penalties should be imposed upon those members who are caught red handed deliberately corrupting the RIPE data base? Is the present RIPE policy of "forgive and forget" with respect to such travesties a product of anyone's intentional design, or is it instead merely the result of utter apathy and indifference on the part of the entire RIPE community? In either case, I believe it to be self evident that the time is... ummm... blooming, blossoming, burgeoning, flourishing, and flowering for this policy to be reexamined and revised. From where I am sitting it is self evident that the current policy of turning a blind eye to these events, and to the bad actors behind them is quite clearly encouraging others to follow suit and to themselves undertake the exact same sorts of travesties. Why shouldn't they? There is no downside. At the very worst, one is simply made to give back all of the stuff that one has stolen, and one is then allowed to go on one's merry way. Regards, rfg P.S. As incensed as I am at the fact that the above named bad actors have been allowed not only to retain their RIPE memberships, but also, apparently, 100% of their legitimately-allocated number resources, this is not even nearly as utterly appalling and inexplicable as the fact that two of the bad actors with direct and provable connections to the prior instances of hijacking that I have reported on publicly this summer have been allowed to remain as officially recognized RIPE IP brokers: https://www.ripe.net/manage-ips-and-asns/resource-transfers-and-mergers/bro… I am referring here specifically to Ebonyhorizon Telecomunicacoes Lda (aka Bitcanal) and also to NetAssist LLC (AS29632), which conveniently continues to maintain its association with, and peering arrangements with both AS57166 aka D2 International Investment Ukraine, Ltd. and, indirectly via D2, AS205869 aka Universal IP Solution Corp., both of which apparently continue to this day to try their best to hijack, either directly or indirectly, via AS Path fraud, a number of IPv4 blocks that clearly do not belong to any of these three Ukranian entities: https://bgp.he.net/AS57166#_peers https://bgp.he.net/AS205869#_peers https://bgp.he.net/AS29632#_peers What IP blocks are these two officially recognized RIPE IP brokers, Ebony Horizon and NetAssist brokering, exactly? Are they blocks that have been successfully hijacked? Does either RIPE or RIPE NCC even know? Does either RIPE or RIPE NCC even care? (One cannot help but wonder what might occur if RIPE were put in charge of distributing meat supplies within Europe. Would RIPE officially designate the McDonald's "Hamburglar" as RIPE's officially recognized agent for said distributions?) I guess that, in the end, all of the questions I have raised above can be boiled down to just one simple question: Who exactly does one need to either kill or maim or seriously wound in order to get kicked out of this organization (RIPE)? Regards, rfg P.P.S. Among the set of six companies / ASNs listed toward the top of this message as possible facilitators of the current hijacks of MNT-SERVERSGET, three are already fairly well known... at least in anti-spam circles... as being among "the usual suspects" when it comes to facilitating spamming and spammers. And no, I do not care to publicly specify which three. It may be true, as the saying goes that "On the Internet, nobody knows that you're a dog", but memories are long, and people don't forget if your company has been repeatedly caught acting like one.

3 4