CANET.db

Just to clarify something Dale said: 
the radbserver distribution he mentioned is as it stands
is running with the old ripe81 database (with all objects in one db file).
It does recognize the route object, though.
The two things that need to be changed are:
- support for split db files (now it needs routes and aut-sys in one file per db)
- support for new whois parameters, esp classless thingy choices.
(Of course one doesn't need to use radbserver for whois queries.)

I plan to make it ripe181 compliant next week,
just after getting a ripe181 going here.
I don't expect it will take a lot of work or changes.

   - r

--------

> From:    "Dale S. Johnson" <dsj at merit.edu>
> To:      config at CAnet.ca
> CC:      rr-impl at merit.edu

> Eric,
> 
> > (Dale: I am still looking forward to beta-test code. I am keeping up
> > a parallel config system to a 181 based system waiting for your release).
> 
> Here's what we have:
> 
>   1) A probably-production-quality "radbserver" which integrates nicely
>      with the new RIPE release.  It requires Perl-5.  Look in
>      ftp.ra.net:pub/radb/tools .
> 
>   2) A policy expression evaluation program "peval" in Gnu C++ and
>      Bison (particular recent revisions are required).  This binary is
>      required for higher level tools.  I believe it works completely,
>      though we are still doing a lot of changes to it.  I'd call this
>      an alpha (still got debugs, etc).  Cengiz at isi.edu and/or
>      ala at merit.edu can point you to it.
> 
>   3) A set of "pol2filters" routines in perl.  This is a rather nice
>      set of routines which read 181 objects, call peval (above), and
>      output net lists with metrics.  The logic is all (?) there for
>      input lists and for route servers.  Dale is trying to rework
>      this code a bit for production usage now.
> 
> (Cengiz also has a "view generator" program which produces special
> Route Server View objects in preparation for config generation).
> 
> If you'd like to get going, I'd recommend bringing up the radbserver
> first, checking it out a bit, and then moving on to peval.  And/or:
> you can run the tools against our server.
> 
> 
> > We should have RIPE data available for bulk transfer shortly (routes are in,
> > crcd.canet.ca is queryable, policy not yet complete). We would like to know
> > how to set up a mnt-by attribute with the RA (you could have told me
> > already; if so please bear with me and tell me again).  If documentation
> > exists, a pointer is fine. Source is "CANET", mnt-by is "CANET-RC".
> 
> Try submitting it to auto-dbm at radb.ra.net.  We have not switched to the
> new 181 software release yet, so the security stuff is not yet there, but
> I think we will accept the data just fine.  We'll probably turn on the
> full 181 release code next week.
> 
> Don't you want your maintainer object to be in CANET.db rather than RADB.db,
> though?  Actually, the whole business of integrating multiple .db files
> is pretty much unexplored so far.  I think this is to be one of the big
> topics at the IETF rr-impl meeting.  I haven't looked at the 181 .db file
> structure yet:  I've heard it is extremely different from 81 (broken into
> separate files for different types of objects).
> 
> Let us know the exact details as soon as your data is available.  We'd
> certainly like to start ftp'ing it each morning, and making it available
> via our server & tools.  I expect Marten is ready to do this, too.
> 
> --Dale
> 
> 
> ===========================================================================
> 
> > From config at enfm.utcc.utoronto.ca Tue Nov 22 19:58:35 1994
> > To: nsfnet-admin at merit.edu
> > From: "CA*NET Routing Coordination" <config at CAnet.ca>
> > Subject: Bulk Reconfiguration of CA*net
> > Cc: rw at mci.net, as3561 at merit.edu, config at CAnet.ca, dsj at merit.edu
> > 
> > Hi,
> > 
> > We would appreciate it if you could append to all CA*net routes with 577
> > anywhere in the announcement list a trailing low priority (usually 4th and
> > 5th priority) 3561(218) 3561(11), subject to MCI's ok that these last two
> > exit points are correct (I know the first is, I tool ). We need this change
> > to be made for the Friday reconfiguration. If it will not make that
> > reconfiguration window, please contact Eric Carroll at eric at canet.ca.
> > 
> > We are setting up for the transition off 690, which we hope to complete by
> > the end of the month. 
> > 
> > We should have RIPE data available for bulk transfer shortly (routes are in,
> > crcd.canet.ca is queryable, policy not yet complete). We would like to know
> > how to set up a mnt-by attribute with the RA (you could have told me
> > already; if so please bear with me and tell me again).  If documentation
> > exists, a pointer is fine. Source is "CANET", mnt-by is "CANET-RC".
> > 
> > (Dale: I am still looking forward to beta-test code. I am keeping up
> > a parallel config system to a 181 based system waiting for your release).
> > 
> > Thanks,
> > 
> > 	Eric Carroll for
> > 	CA*NET Routing Coordination
> > 



-------- Logged at  Wed Nov 23 17:04:12 MET 1994  ---------