Subject: Re: nore on disk stats
To: None <perry@piermont.com>
From: Charles Hannum <Charles-Hannum@deshaw.com>
List: tech-kern
Date: 11/14/1995 17:50:24
   Charles Hannum writes:
   >    Ick. I wouldn't mind having netstat use sysctl(), but I think that
   >    going for SNMP is REALLY BAD here.
   > 
   > So it could use the local interface if a host name isn't specified.
   > This is a detail.

   Its not a detail. The day I see an ASN.1 compiler being used in the
   NetBSD kernel build process (probably doubling its size) I move to
   some other operating system.

We're not talking about putting ASN.1 into the kernel.  I said
`SNMP-like', meaning it takes the same sort of structured queries and
the MIB is arranged in a tree structure.  sysctl() already does this.
What it doesn't do is: 1) supply type information, 2) have iteration
primitives, 3) have enough things implemented, and 4) conform to an
existing layout.  These are fairly easy to correct.  All the ASN.1
bloat (which is part of the network protocol, not the basic
functionality) can go in snmpd and possibly a SNMP shared library.

   Well, actually, there are several perfectly good snmpd's we could be
   running -- both the gigantic ISODE snmpd and the (by comparison)
   smaller CMU snmpd have been ported to BSDI at various points (and the
   ISODE one actually used to refer to itself as the 4.4 SNMPd in its man
   pages) so they should run without any problem if we wanted an snmpd.

The point of this exercise, and the reason this discussion was
started, is to get rid of klugy interfaces that rely on groveling
through the kernel.

   All things considered, this is a lot of design and implementation work
   to solve a non-problem with a really ugly solution.

The design already exists.  The implementation is fairly trivial.  The
problem is real.  The solution reuses existing standards.  And, it
doesn't require having two complete MIBs.