Re: [P2PSIP] [OPSAWG] An SNMP Usage for RELOAD

Just to make sharper the statements made by Tom. In the SNMP architecture in order for two entities to exchange management information we have a protocol (basic primitives to fetch pieces of information, set values of pieces of information or send notifications related to the pieces of information) and a data model (e.g. MIB model) expressed in a Data Modeling Language (e.g. SMI). The proposed text talks about the protocol (SNMP) but not about the information exchanged (data model). This is of little help. 

Dan

> -----Original Message-----
> From: opsawg-bounces@ietf.org [mailto:opsawg-bounces@ietf.org] On
> Behalf Of t.petch
> Sent: Thursday, October 27, 2011 12:09 PM
> To: peng.yonglin@zte.com.cn
> Cc: hao.zhenwu@zte.com.cn; opsawg@ietf.org; p2psip@ietf.org
> Subject: Re: [OPSAWG] An SNMP Usage for RELOAD
> 
> ----- Original Message -----
> From: <peng.yonglin@zte.com.cn>
> To: "t.petch" <ietfc@btconnect.com>
> Cc: <hao.zhenwu@zte.com.cn>; <opsawg@ietf.org>; <p2psip@ietf.org>;
> <wang.wei108@zte.com.cn>; <meng.yu@zte.com.cn>
> Sent: Wednesday, October 26, 2011 5:02 AM
> 
> > In short, the relation between SNMP Usage and RELOAD as follows:
> > SNMP Usage get a available IP address, which is used for SNMP
> > communication, for snmeEngineID or NodeID or ResourceID or Manager
> name by
> > RELOAD mechanism, such as AppAttach, Store, Fetch.
> > Then SNMP Usage uses this IP address for SNMP communication between
> SNMP
> > entities by the SNMP mechanism and architecture which defined in
> RFC3411
> > and RFC5953 and other SNMP RFCs (as depicted in the diagram below).
> And
> > now there is little relation between SNMP Usage and RELOAD besides
> using
> > RELOAD certificate. So we say "SNMP entities talk to each other using
> SNMP
> > protocol on dedicated connections"
> 
> No, it still does not make sense.  You understand that an SNMP engine
> can do
> nothing until it has been customised - snmpEngineID,
> snmpTargetAddrTable, etc -
> and you use RELOAD to obtain that information; I assume you are
> familiar with
> the MIB modules into which that information must then be placed before
> an engine
> can communicate.
> 
> Two engines can then communicate but cannot do anything useful without
> suitable
> MIB modules; the I-D says
> " SNMP usage for RELOAD SHOULD provide the management functions ...
>      ... monitoring the number of the messages
>    initiated, forwarded or processed by nodes, reporting program
> failure
>    , message forwarding failure or other error on nodes."
> 
> This is a typical usage for SNMP but is quite impossible unless and
> until a
> suitable
> MIB module is defined, which is what everyone using SNMP for management
> then does.  Your earlier response to David rejected the idea of
> creating a MIB
> module, which leaves me with two SNMP engines able to talk to each
> other but
> having nothing useful to say.
> 
> The I-D also says
> "As traditional network
>    management protocols (e.g., SNMP) cannot be directly applied to
>    RELOAD network management, it is necessary to introduce new RELOAD
>    usage of SNMP."
> which is only true in the sense that SNMP cannot be directly applied to
> any
> network
> management, unless and until there is a suitable MIB module.  Once
> there is,
> and you have two communicating engines (which you say you will have),
> then
> of course SNMP can be used to manage anything, including RELOAD.
> 
> Tom Petch
> 
> >
> > If you till have questions, please tell us. Thanks!
> >
> >
> > *************************************************
> > 邮 件：peng.yonglin@zte.com.cn
> > 内 线：81543
> > 外 线：025-52871543
> > 手 机：13776637274
> > 传 真：025-52872187
> > *************************************************
> >
> > >
> >
> >
> > > ----- Original Message -----
> > From: <peng.yonglin@zte.com.cn>; <peng.yonglin@zte.com.cn>
> > To: <p2psip@ietf.org>; <opsawg@ietf.org>; <p2psip@ietf.org>;
> > <opsawg@ietf.org>
> > Cc: <hao.zhenwu@zte.com.cn>; <hao.zhenwu@zte.com.cn>
> > Sent: Tuesday, October 25, 2011 4:23 AM
> >
> > > We have done more research on security and other issues based on
> David's
> > > comments, and updated our draft.
> > >
> > > This is the new version of draft:
> > > http://www.ietf.org/id/draft-peng-p2psip-snmp-03.txt
> > >
> > > We welcome any comments! Thanks!
> > >
> > > Our analysis of David's comments as follows(blue text):
> > >
> > > 1) Use SNMPv3 terminology
> > > Apparently, you come from the RELOAD side of things, not the SNMP
> > > side.
> > > Your text doesn't use the common SNMP terminology for various
> > > SNMP-related things.
> > > You will have a better chance of success if you describe SNMP-
> related
> > > things using the normal SNMPv3 terminology.
> > > This isn't critical; your ideas seem reasonable, but since you
> don't
> > > use the standard terminology, you might mean slightly different
> things
> > > than what it would mean if you used the standard terminology.
> > >
> > > (I don't have a RELOAD background, so some of my comments might
> seem
> > > wrong because I don't know the RELOAD concepts and terminology. Now
> > > you'll understand my point ;-)
> > >
> > > PYL: We have updated the draft using SNMP terminology. As we are
> not
> > SNMP
> > > experts, some terminology may not be accurate. We will try to
> improve
> > this
> > > over time.
> >
> > You are indeed using SNMP terminology, but I struggle to see how its
> usage
> > is
> > the same as in RFC3411.
> >
> > RELOAD defines a messaging network, on top of which run applications,
> > which
> > RELOAD calls Usages, so when you define an SNMP Usage, you would
> appear to
> > be
> > discarding most of an Architecture for SNMP as defined in RFC3411,
> perhaps
> > providing instead a gateway for functionality such as that specified
> in
> > RFC3414;
> > or perhaps not, I do not find the I-D clear on this.  Thus you have
> >
> >                   Application
> >
> >             +-------+  +-------+  +-------+
> >             | SIP   |  | XMPP  |  | SNMP  |  ...
> >             | Usage |  | Usage |  | Usage |
> >             +-------+  +-------+  +-------+
> >          ------------------------------------------- Messaging API
> >
> >                       RELOAD
> >
> > but elsewhere you say
> > ' SNMP entities talk to each other using SNMP protocol on dedicated
> > connections'
> > which is rather different.
> >
> > The I-D reads more like a wish list of requirements, than an
> overview,
> > architecture or framework.  Or perhaps it is all beyond me:-(
> >
> > Tom Petch
> >
> > <snip>
> 
> _______________________________________________
> OPSAWG mailing list
> OPSAWG@ietf.org
> https://www.ietf.org/mailman/listinfo/opsawg