- Test - lists.nanog.org

Re: Trouble Ticket Systems
by Robert E. Seastrom 04 May '95

04 May '95

MeLiSe <melise(a)BBN.COM> said: > > The Remedy Action Request System (ARS) is flexible (can build it > with several different dbs as underpinning), economical and their > support desk is very helpful. As if it needed any more endorsement, ARS rocks.

1 0

Re: Trouble Ticket Systems
by MeLiSe 04 May '95

04 May '95

Eric, The Remedy Action Request System (ARS) is flexible (can build it with several different dbs as underpinning), economical and their support desk is very helpful. Melise Jones BBN STD

1 0

Re: PRDB retirement (and note about AS690 advisories)
by Dale S. Johnson 04 May '95

04 May '95

Alan, > I also want to echo Sean Doran's question of what text to put at the top > of any NACR submitted before 8 May. It's not just the text at the top > that presents a problem; even the "%begin nsfnet nacr v7.1" or "%begin > ansnet nacr v2.0" no longer seems particularly apppropriate, and what do > we put in the "aup:" line? To reiterate Brian's answer to this: We no longer care what text is at the top. Feel free to use aup: N ; nobody will mind any more. This data will be discarded next week. Please do not mess with the %begin lines on the NACR, or your NACR will definitely be mishandled. The PRDB is something of an old dog at this point, and we'd rather not try to make him learn new tricks for his final few days. --Dale

1 0

Re: PRDB Retirement next Monday, May 8
by Dale S. Johnson 03 May '95

03 May '95

Sean, > >The one question I have about NACRs until 8 May is what > >text to put at the top of them. > > > >Obviously the old NSFNET wording is no longer appropriate. > > > > Sean. > > My feeling on this is that ANS doesn't look at it, so you can > replace it with XXXXXX if you like, or not if you don't, but I will > let Merit have the last word on this. We have disabled any checking of this. You leave the NSFNET AUP text as for sentimental value, or you may delete it, or you may replace it with anything that you like. (Though actually our preference is for tasteful jokes or poems ;-). --Dale

1 0

Re: PRDB Retirement next Monday, May 8
by Steve Heimlich 03 May '95

03 May '95

Sean, >The one question I have about NACRs until 8 May is what >text to put at the top of them. > >Obviously the old NSFNET wording is no longer appropriate. > > Sean. My feeling on this is that ANS doesn't look at it, so you can replace it with XXXXXX if you like, or not if you don't, but I will let Merit have the last word on this. Steve

1 0

PRDB retirement (and note about AS690 advisories)
by Steve Heimlich 03 May '95

03 May '95

Folks, Thanks to Dale for the initial update. Let me add a further note on our transition of AS690 from NSFnet routing policy toward ANS routing policy... Shortly beyond retirement of the PRDB, ANS will halt our use of AS690 advisories. We hope to do this soon after retiring the PRDB, using a tool which converts existing announcements into policy expressable using "pure" RIPE-181; this resulting aut-num object will work with various standard RIPE/PRIDE tools. After that, we'll be culling our very large aut-num object into something smaller which is uses AS-based policy as much as possible, with a minimum of prefix-based exceptions. New registered prefixes will assume the current majority policy toward the home AS in which they're registered. The size of the aut-num object right now comes from handling the large number of exceptions for prefixes originating within a given AS (which we're using so that we disturb actual routing as little as possible initially). The configuration of metric:as lists has been a challenge to manage to say the least, particularly through the transition. As they are obviously burdensome for ANS as well, we have a mutual interest in witnessing their retirement, and we're working to do this as quickly as possible. In addition, info in the RADB which has as its source "PRDB" can be ignored on a per-provider basis, so that eventually the PRDB-derived info can go away entirely. So, if provider X tells us to prefer source:X over source:PRDB, we can work out migration details so the old PRDB info can be knocked out in one action, replaced by the preferrable RIPE-181 source. Steve --------------- Dale's original follows ---------------- What Networks Need to Do When the PRDB Is Retired On Monday, May 8, the Policy Routing Database (PRDB) will be turned off. The PRDB has been used to configure the NSFNET Backbone Service and ANSNET (AS690) since 1989. After May 8, configurations for the ANSNET backbone will be generated from the Routing Arbiter Database (RADB). The PRDB and the RADB have been running with parallel data for over a year. As of May 8, the last data from the PRDB will be transferred to the RADB, and the PRDB will be permanently retired. FROM NACRS TO ROUTE OBJECTS Starting 09:00 EST May 8, NACRs will no longer be accepted for AS690 configurations; the way to submit additions or changes of route announcements for AS690 will be to submit Route objects by e-mail to auto-dbm(a)ra.net. If you submit a NACR after 09:00 May 8, you will receive a reply that gives you a translation of that NACR into a Route object and explains how to mail that Route object to auto-dbm(a)ra.net for immediate inclusion in the RADB. For an overview of the RADB and more information about creating and submitting Route objects, see: http://www.ra.net (select Routing Arbiter from the Projects menu) ftp://ftp.ra.net/routing.arbiter/radb/OVERVIEW ftp://ftp.ra.net/routing.arbiter/radb/doc/how_to_register WILL YOU BE AUTHORIZED TO UPDATE THE RADB AFTER MAY 8? Under the old NACR/PRDB system, each AS maintained a list of "valid requestors" who could make routing changes for that AS. Under the new Route object/RADB system, each network has an Origin AS ("Home AS"); the individuals who can make routing changes for that AS's networks are listed in the Maintainer object for that AS. Initial versions of Maintainer objects were created from the PRDB "valid requestor" information on February 1, and many more Maintainer objects have been added to the RADB since then. Thus it is very likely that your authorization in the RADB has already been established. If you have successfully submitted a NACR since February 1, then you are already authorized to submit Route objects for that AS. Note, however, that if another organization has been submitting NACRs for you, that organization may no longer be authorized to do this under the RADB unless you specifically include e-mail addresses for that organization in your Maintainer object. Here is the breakdown of authorized submitters: Date Who Can Make Routing Changes ------ ------------------------------ Before Feb. 1 Any valid requester of any AS in the network's announcements ("aslist"), including the Primary AS Feb. 1 - May 8 Persons specified by the Maintainer objects of *either* the network's Primary AS or its Home AS After May 8 *Only* persons specified by the Maintainer object of the network's Home AS The e-mail addresses for people authorized to make changes for your network must be listed on separate MAIL-FROM lines in your Maintainer object. (A more elaborate system can be used after May 8.) To verify that your Maintainer object contains the correct information, enter a command such as the following: whois -h whois.ra.net MAINT-AS690 For more information on checking, submitting or updating Maintainer objects, see: http://www.ra.net (select Routing Arbiter from the Projects menu) ftp://ftp.ra.net/routing.arbiter/radb/doc/updating_maintainers If you have further questions about registering in the RADB, send e-mail to DB-admin(a)ra.net. --Dale Johnson Merit/RA

2 3

Re: PRDB Retirement next Monday, May 8
by Sean Doran 03 May '95

03 May '95

The one question I have about NACRs until 8 May is what text to put at the top of them. Obviously the old NSFNET wording is no longer appropriate. Sean.

1 0

Re: PRDB Retirement next Monday, May 8
by M.H.Behringer＠dante.org.uk 03 May '95

03 May '95

My understanding of the situation for Europe is (and I'm sure Merit or ANS or somebody else will correct me if I'm wrong): - Currently we (Europe) have to register in the RIPE DB and send NACRs - As of next monday... ... we won't need to send NACRs any more. (nor something to the RADB) ... Merit will mirror the RIPE DB into the RADB, so that the RADB contains the RIPE data as well. This will be done on a, say, daily basis. (that was the assumption, right?) ... ANS (and others) access the RADB, which contains all information - At a later stage the whole system should be fully distributed, as discussed several times. Now, the one possible problem I can think of is that Merit is not able to do the mirroring until next Monday. In that case we (Europe) would have to send the route objects to the two DBes (RIPE and RADB) for a certain time to be defined. This would be not optimal, but it would work. Still I very much hope that we don't have to go for this. Comments? Michael

2 1

Re: Topics for the May 22-23 NANOG
by Scott Bradner 03 May '95

03 May '95

> How about resurecting the idea of reviewing RFC's and drafts for operational impact? as one of the OPS ADs I'd welcome anyone who would be willing to help in this task Scott

2 1

Re: PRDB Retirement next Monday, May 8
by Hank Nussbacher 03 May '95

03 May '95

On Mon, 1 May 1995 20:20:22 -0400 you said: >WILL YOU BE AUTHORIZED TO UPDATE THE RADB AFTER MAY 8? > >Under the old NACR/PRDB system, each AS maintained a list of "valid >requestors" who could make routing changes for that AS. Under the new >Route object/RADB system, each network has an Origin AS ("Home AS"); the >individuals who can make routing changes for that AS's networks are >listed in the Maintainer object for that AS. Initial versions of >Maintainer objects were created from the PRDB "valid requestor" >information on February 1, and many more Maintainer objects have been >added to the RADB since then. Thus it is very likely that your >authorization in the RADB has already been established. If you have >successfully submitted a NACR since February 1, then you are already >authorized to submit Route objects for that AS. The RADB has my entry as follows: mntner: MAINT-AS378 descr: ILAN admin-c: Hank Nussbacher tech-c: Hank Nussbacher upd-to: hank(a)vm.tau.ac.il auth: MAIL-FROM HANK(a)taunivm.tau.ac.il mnt-by: MAINT-AS378 changed: DB-admin(a)ra.net 940302 source: RADB The RIPE DB has my entry as follows: mntner: AS378-MNT descr: ILAN upd-to: hank(a)vm.tau.ac.il mnt-nfy: hank(a)vm.tau.ac.il auth: CRYPT-PW somecryptstring mnt-by: AS378-MNT admin-c: Hank Nussbacher tech-c: Hank Nussbacher changed: hank(a)vm.tau.ac.il 941122 source: RIPE I would assume that everyone in Europe has updated their mntner object and customized as they please. Now we have to do it again? Couldn't the RADB just copy all the mntner objects as listed in RIPE? > --Dale Johnson > Merit/RA Thanks, Hank Nussbacher Israel

2 1