Looks like PSInet is demonstrating their complete and total lack of clue
again.
Who has less of a clue? A provider that announces a 1.1.1.0/24, or
a provider that listens to junk like 1.1.1.0/24?
It doesn't matter what junk is announced, *IF* you could get providers
not to listen to it.
On Wed, May 13, 1998 at 05:07:38PM -0500, Sean Donelan put this into my mailbox:
>Looks like PSInet is demonstrating their complete and total lack of clue
>again.Who has less of a clue? A provider that announces a 1.1.1.0/24, or
a provider that listens to junk like 1.1.1.0/24?It doesn't matter what junk is announced, *IF* you could get providers
not to listen to it.
This would get into the discussion about authenticating BGP again,
but what happens if IANA and ARIN get all cozy (no, I don't want to
think about it) and ARIN decides to allocate some /18's out of 1/8?
How do the folks filtering this network (because it's a reserved number)
know that it's gone from being reserved to active use?
Maybe i'm just nitpicking, but with the so-called 'IP shortage', it
would make some amount of sense that some of these reserved blocks will
be opened up more and more. I don't like the idea of screwing over
some poor new guy's connectivity just because engineering folks have
better things to do than do 'whois's on their filters every week.
Or is there some historical/technical reason that I'm not aware of
for not listening to 1/8?
-dalvenjah
This would get into the discussion about authenticating BGP again,
but what happens if IANA and ARIN get all cozy
They already are all cozy, i.e. ARIN allocates address blocks delegated to
it by IANA and if you want to appeal an ARIN decision, the appeal goes to
IANA. Chances are that Magaziner's white paper due out today will suggest
formalizing that relationship.
(no, I don't want to
think about it) and ARIN decides to allocate some /18's out of 1/8?
How do the folks filtering this network (because it's a reserved number)
know that it's gone from being reserved to active use?
One item that is currently before the ARIN Advisory Council is a
suggestion that we publish a map of all currently allocated IP space and
keep that map up to date. If we do this then I think it solves this
problem if people keep verifying their filters against the map or if they
generate their filters based on the map.
Maybe i'm just nitpicking, but with the so-called 'IP shortage', it
would make some amount of sense that some of these reserved blocks will
be opened up more and more. I don't like the idea of screwing over
some poor new guy's connectivity just because engineering folks have
better things to do than do 'whois's on their filters every week.
Rather than whining about the "powers that be" and pointing out how the
system is broken, why not propose a solution that these so-called "powers"
could implement to solve the problem. IANA and ARIN and RIPE and APNIC are
composed of some very ordinary yet clueful Internet folks that are capable
of recognizing a good idea when it is explained to them.
The ARIN AC is having a face-to-face day of meetings at the end of June
followed by a Board of Trustees meeting the following day. If there is
stuff that we can fix or we can do better, tell us all about it soon so we
can get it on the agenda for that meeting.
So what are you saying here Sean? We don't have a clue? 3/4 of the
Internet service providers must not have a clue then. I don't belive I
have ever read anywhere that 1/8 or anything smaller wasn't a valid
route.
Maybe you would like to show what you have in your BGP tables for 12/8?
How about 9/8?
How about 24/8?
How about 32/8?
How about 44/8?
Do you want me to go on??
T..S
OK, Good Idea People...here's one.
The ARIN WhoIs database contains many false entries, as does InterNIC's.
The proposal made to InterNIC (which was ignored in so many words) was that
when we users notice a bogus entry, we notify the registrar who contacts
the registration owner for a correction or, if it's a blatant
falsification, deletes the entry.
This is a reasonably painfree way to properly administer a database relied
upon by the entire world and just happens to comply with several RFCs.
As an example of one I reported last week to ARIN (who suggested I contact
my ISP for the info I'm after...how clueless):
Lifenet Imarketing Group, Inc. (NETBLK-SPRINT-D01D98)
3090 Pullman Street, suite B
Costa Mesa, CA 92626
US
Netname: SPRINT-D01D98
Netblock: 208.29.152.0 - 208.29.152.255
Coordinator:
Warmington, Drew (DW5183-ARIN) nomailbox@NOWHERE
7149661345 (FAX) 714-263-9260
Kindly note that the email address doesn't even have a TLD. Hard to
consider it valid, what?
As a side note, why not modify your software so that things like this are
rejected? It shouldn't be all that hard to make sure that the email
address at least has the required components.
Spam(tm) is pressed meat. Spammers should be too.
Dean Robb
PC-Easy
On-site computer services
(757) 495-EASY [3279]
I have yet to see a mandate that stated that we must filter any prefix
longer than /19. If you choose to filter at /19 that is your choice. We
choose to have connectivity that isn't hampered to someone who actually
has a classfull C somewhere on the net. As for filtering 1.1.1.1/24 or
the entire block because it is reserved for that matter, who is to say
that with the address space thinning out that the IANA won't start
allocating it? Some poor provider is going to be route filtered on a
totally legitimate block if we use your reasoning.
Filtering announcements from your downstreams for blocks that they are not
allocated on the other hand IS not only legitimate but something that I
would consider a moral responsibility of every provider. If they are not
multi-homed, why listen to ANYTHING that comes from them? Static route
them and you have no worries.
So, in answer to your question, it is the provider who makes the bogus
announcement for whatever reason they made it and not the providers who
listen to it who has less of a clue.