comp.os.linux.networking
[Top] [All Lists]

Re: Subnetting/Routing

Subject: Re: Subnetting/Routing
From: Moe Trin
Date: Sat, 14 May 2005 20:40:58 -0500
Newsgroups: comp.os.linux.networking
In article <kxehe.1450$sb5.1104@xxxxxxxxxxxxxxxxxxxxxxxxxx>, Wolf wrote:

>You have not been doing your homework. Have you actually tried to get 253
>netmask using "the bits?" You can't get there.

You are not bothering to read the documents that were pointed out for you
that refute your incorrect interpretations.  Please read the official
standards.   Start here:

2026 The Internet Standards Process -- Revision 3. S. Bradner. October
     1996. (Format: TXT=86731 bytes) (Obsoletes RFC1602) (Updated by
     RFC3667, RFC3668, RFC3932, RFC3979, RFC3978) (Also BCP0009) (Status:
     BEST CURRENT PRACTICE)

particularly sections 3 and 4 of that document. Then read the two standards
applicable to subnet masks:

0950 Internet Standard Subnetting Procedure. J.C. Mogul, J. Postel.
     Aug-01-1985. (Format: TXT=37985 bytes) (Updates RFC0792) (Also
     STD0005) (Status: STANDARD)

1122 Requirements for Internet Hosts - Communication Layers. R.
     Braden, Ed.. October 1989. (Format: TXT=295992 bytes) (Updated by
     RFC1349) (Also STD0003) (Status: STANDARD)

Look up thread if you want to learn the specific sections to read if you
are in a hurry. Then read the _proposed_ standard for CIDR (can you guess
why that word is underlined?)

1519 Classless Inter-Domain Routing (CIDR): an Address Assignment and
     Aggregation Strategy. V. Fuller, T. Li, J. Yu, K. Varadhan. September
     1993. (Format: TXT=59998 bytes) (Obsoletes RFC1338) (Status: PROPOSED
     STANDARD)

and read section 4.1 of that document.

>This should help:  Take out a sheet of paper and start by creating 8
>columns.

and fix your browser (or use a real news client) so that you are not
using a proportional font when trying to show tables.  I know it may come
as a surprise to you, but those of us who have been network admins for more
that five or ten years really do have that table memorized - in CIDR form,
decimal and hex masks, and the number of hosts in the block.

>Discontiguous networks are ugly and a sign you did not do your work right.

Your opinion is unimportant, but they are also not recommended by RFC1219,

1219 On the assignment of subnet numbers. P.F. Tsuchiya. Apr-01-1991.
     (Format: TXT=30609 bytes) (Status: INFORMATIONAL)

but if you read RFC2026 you would discover that the _standard_ does permit
them.  Also, if you read the post you replied to, you would discover I am
quite aware that few O/S will tolerate such masks. I don't know anyone still
using MIT's concept (RFC0932), but that is one example of a working setup
where they _required_ non-contiguous bits in the subnet mask. Jon Postal's
version (RFC0925) didn't _care_ about the mask.

If you think that the whole world must use the slash notation to express
network masks, how do you explain the assignments by the five RIR
(AFRINIC|APNIC|ARIN|LACNIC|RIPE) and why they use CIDR notation as a
secondary means of indicating network size. Could it be because they
are not assigning netblocks in binary sizes??? Hmmm, what is the mask for
a network of 768 hosts? There are 3 such assignments in AFRINIC, 416 in
ARIN, and 52 in RIPE.  How about one for 1536 hosts? There are 2 such
assignments in AFRINIC, 210 in ARIN, and 27 in RIPE. In fact, the latest
copy of the RIR zone files I have has 201 different sizes of allocations.

>I have a really good book on the subject from a class I took last year:

I started in IP network administration in the spring of 1986, but had
access back when most of the net that spoke TCP was using a ARPA address
in the 10.0.0.0/8 range.  You?

        Old guy

<Prev in Thread] Current Thread [Next in Thread>