[db-wg] Cleanup of route(6) objects in the RIPE NONAUTH database using unregistered space

For what it's worth, this still has my support.

-Cynthia

On Fri, May 7, 2021 at 12:56 PM denis walker <ripedenis at gmail.com> wrote:
>
> HI Ed
>
> As no one has objected I think we can assume you can go ahead with this plan.
>
> cheers
> denis
> co-chair DB-WG
>
> On Fri, 7 May 2021 at 09:20, Edward Shryane <eshryane at ripe.net> wrote:
> >
> > Hi Cynthia, Denis,
> >
> > Thanks for your feedback. I will plan to implement a cleanup job without an NWI, and will keep the DB-WG informed.
> >
> > Regards
> > Ed Shryane
> > RIPE NCC
> >
> > On 25 Apr 2021, at 13:03, Cynthia Revström <me at cynthia.re> wrote:
> >
> > imo an NWI seems unnecessary unless we want it for purposes of referencing back to in the future.
> >
> > -Cynthia
> >
> > On Wed, Apr 21, 2021, 04:02 denis walker via db-wg <db-wg at ripe.net> wrote:
> >>
> >> Colleagues
> >>
> >> How do you want to proceed with this suggestion? Do we have support
> >> for deleting these ROUTE(6) objects for deregistered space and where
> >> the objects were created before registration? Does anyone have any
> >> objections to this as an ongoing cleanup?
> >>
> >> Do we even need an NWI or can we just ask the NCC to go ahead with it?
> >>
> >> cheers
> >> denis
> >> co-chair DB-WG
> >>
> >> On Thu, 11 Mar 2021 at 15:07, Edward Shryane via db-wg <db-wg at ripe.net> wrote:
> >> >
> >> > Dear colleagues,
> >> >
> >> > In response to the recent discussion on "196.52.0.0/14 revoked, cleanup efforts needed", I'd like to propose a regular automated cleanup of route(6) objects in the RIPE-NONAUTH database using unregistered space.
> >> >
> >> > Regards
> >> > Ed Shryane
> >> > RIPE NCC
> >> >
> >> >
> >> >
> >> > Problem Definition
> >> > ------------------
> >> > When an RIR deregisters IPv4/IPv6 address space, any route(6) objects in the RIPE-NONAUTH database using that address space are not cleaned up (deleted).
> >> >
> >> >
> >> > Solution Definition
> >> > -------------------
> >> > Once a day, compare route(6) IPv4/IPv6 prefixes in the RIPE-NONAUTH database against the combined delegated stats from all RIRs.
> >> >
> >> > Only route(6) prefixes that are "allocated" or "assigned" in any RIRs delegated stats should remain in the RIPE NONAUTH database. If a prefix is "available" or "reserved" then it is considered to be unregistered, and any associated route(6) objects will be eligible for deletion.
> >> >
> >> > If a prefix is partially "available" or "reserved" then it is also considered to be unregistered.
> >> >
> >> > If a prefix is not listed in any RIRs delegated stats, that prefix is skipped.
> >> >
> >> > The origin AS status is not considered, only the IPv4/IPv6 prefix.
> >> >
> >> > If a newly unregistered prefix is discovered, first allow a grace period of 1 week. This allows time for mistakes in the delegated stats to be corrected.
> >> >
> >> > After 1 week, contact the route(6) maintainer(s) to notify them that the route(6) object will be deleted.
> >> >
> >> > After a further 2 weeks, delete the route(6) object.
> >> >
> >> > A maintainer can request the RIPE NCC to exclude a route(6) prefix from deletion (for example, the deregistration of the prefix is being disputed). For any excluded prefixes, any associated route(6) objects will not be deleted.
> >> >
> >> >
> >> > Impact Analysis
> >> > ------------------
> >> > There are approximately 738 routes (out of 56,230) in the RIPE-NONAUTH database using a prefix with status "available" or "reserved" in an RIRs delegated stats, which will be eligible for deletion.
> >> >
> >> > There are approximately 94 route6's (out of 1,564) in the RIPE-NONAUTH database using a prefix with status "available" or "reserved", so also eligible for deletion.
> >> >
> >> > There are approximately 64 routes and 37 route6's with a prefix not listed in any RIR's delegated stats. These will not be affected.
> >> >
> >> > When this cleanup is implemented, the backlog of route(6) objects eligible for deletion will be processed at the same time, leading to a large amount of emails to affected maintainers. Once this backlog is processed, the number of route(6) objects affected is expected to be low.
> >> >
> >> >
> >> >
> >>
> >