Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 27 Oct 1998 11:05:00 -0800
From:      Terry Lambert <terry@whistle.com>
To:        Andrzej Bialecki <abial@nask.pl>
Cc:        small@FreeBSD.ORG
Subject:   Re: Unified Configuration Interface
Message-ID:  <3636195C.535@whistle.com>
References:  <Pine.BSF.4.02A.9810271312340.3398-100000@korin.warman.org.pl>

next in thread | previous in thread | raw e-mail | index | archive | help
[ ... transactional, possible to expose via standard protocol ... ]

Andrzej Bialecki wrote:
> Sounds like a good idea (I mean, the transactional model).
> Also, I would stress the word "possible" in the above statement
> - thus far all implementations of LDAP or SNMP agents I've seen
> are heavy-weight (from my point of view).

Mine too; I wasn't thinking in terms of a server, necessarily,
though, since I could see the following from an embedded system:

1)	SLP request: where is the LDAP server?

2)	LDAP request: where is my configuration data?

3)	<OTHER> request: Please manage me via data changes
	in the LDAP data store, and tell me when changes
	have taken place so I can reconfigure myself.

That would make it most probably a client implementation only.
I mention ACAP because it's a fairly small client implementation
as well.

The main issue is that there needs to be someone, somewhere,
to "be the king"; whether this just means a machine to export
an NIS+ service, or whether this means SNMP or LDAP is undefined.


[ ... List of ?RFC's specifying MIB's ... ]

] Ugh.. Yeah.... This is pretty extensive list. But you make a
] good point...
] 
] What worries me, though, is that the only one (free)
] implementation of snmp agent that I'm aware of is, well, more
] than bulky...

We should make an early distinction here between "schema" and
"implementation".  A MIB specifies elements and hierarchical
relationships between the elements.  You don't have to use
SNMP to be able to use a MIB.  Anything capable of implementing
a hierarchy, including a file system, could be used.  So could
LDAP, ACAP, SNMP, or mySQL.

I think the most important part of the process at this point
is a definition of schema for the things you want to be able
to manage; from personal experience, I have to say that it's
always the thing that falls by the wayside that makes the
resulting code either not work, or work in some standards
non-conformant way that doesn't play well with others.  From
my poerspective, either one of these would be a loss, and
either one of these wold prevent the results from scaling to
either clustes of embedded machines, or large FreeBSD (or
other) boxes.

> > 3)    If the intent is small, standalone servers/devices
> >       (one of the stated intents of the freebsd-small list
> >       is to explore embedded systems, so this is appropriate),
> >       it is probably worth while to consider an implementation
> >       of SLP (Service Location Protocol, RFC 2165) for peer
> >       discovery and integration.
> 
> Thanks for the pointer - I'll look at it.

If you need the reference implementation patches after you
get the reference implementation (from Sun; the email address
is <Charles.Perkins@Eng.Sun.COM>), I can give them to you.
The license says:

	Users may copy or modify Sun Linux SLP without
	charge, but are not authorized to license or
	distribute it to anyone else except as part of
	a product or program developed by the user.

I take this to read that if I had a program I could give you,
that I could give you the code.  I don't, but I believe that
Pico/FreeBSD would qualify in terms of being able to give
the code out to someone else.

> > Also, although currently classed "experimental", consideration
> > should be given to RFC 2307, An Approach for Using LDAP as a
> > Network Information Service, since it addresses most NIS+
> > configuration information, including bootp and machine
> > information.
> 
> I'm not familiar with LDAP that much... Is there any
> implementation of it which takes less than, say, 200kB?

My current SLAPD (LDAP server) is 160k, linked shared against
libc_r and libcrypt, so I think for a static version, the
answer is probably "no" (but it could be crunched, and I
don't know the impact of libc_r vs. libc on a crunched disk).


But the point of mentioning RFC 2307 wasn't to specifically
advocate LDAP; it was to point at a MIB that defines most of
the things needed for storing all NIS+ data.  The MIB is, as
above, schema information, totally seperate from implementation.


-- Terry Lambert
-- Whistle Communications, Inc.
-- terry@whistle.com

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-small" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3636195C.535>