[tkined] Proposed change of the snmp command.

Juergen Schoenwaelder (schoenw@gaertner.de)
Thu, 11 Sep 1997 16:12:00 +0200

The current snmp session commands suffer from semantic parameter
overload. For example, if you create a session and you want to send a
trap, you have to re-configure the session handle before sending the
trap and you have to configure it back afterwards. If you want to
receive traps, you have currently no way to specify the transport
endpoint on which you want to listen for traps.

My first attempt to address this problem was to add additional
configuration options (-trapaddress, -trapport...). This does however
not lead to a nice and simple solution. The number of options
increases quite a bit and the C code is getting confusing because you
have to carefully check which address to use etc.

I am currently thinking of using the SNMPv3 idea where they have split
manager and agents into smaller things called "command generators",
"command responders", "notification originators" and "notification
receivers". Taking this approach would mean that there will be 4
different types of SNMP session handles. The command set supported by
a session handle will be defined by the session type:

A "command generator" handle will support:

configure, cget, destroy,
wait, get, getnext, getbulk, set, walk, scalars, ...,
bind {send|recv|report}

A "command responder" handle will support:

configure, cget, destroy,
bind {send|recv}, instance, info {instances},
bind {get|set|create|check|commit|rollback|begin|end}

(I will probably make instances accessible as Tcl handles
to simplify their handling in Tcl.)

A "notification originator" handle will support:

configure, cget, destroy,
wait, trap, inform, bind {send|recv}

A "notification receiver" handle will support:

configure, cget, destroy,
bind {send|recv|trap|inform}

Before makeing the change in my sources, I would like to hear if there
are any problems with this approach. I think this change makes things
much simpler and cleaner. Note, this change will introduce some
incompatibilities, which should however be easy to fix by introducing
additional session handles. Comments?
Juergen

-- 
Juergen Schoenwaelder  schoenw@ibr.cs.tu-bs.de http://www.cs.tu-bs.de/~schoenw
Technical University Braunschweig, Dept. Operating Systems & Computer Networks
Bueltenweg 74/75, D-38106 Braunschweig, Germany.     (Tel. +49 531 / 391-3283)
--
!! This message is brought to you via the `tkined & scotty' mailing list.
!! Please do not reply to this message to unsubscribe. To subscribe or
!! unsubscribe, send a mail message to <tkined-request@ibr.cs.tu-bs.de>.
!! See http://wwwsnmp.cs.utwente.nl/~schoenw/scotty/ for more information.