| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
I have been quite intriqued by discussion on this issue on the
reflector.
A popular CEO once told me that if you cannot summarize your
product pitch to a lay person in a 1/2 minute then you have
a serious problem of not knowing what you have. We are strugling
with this when it comes to RPR. On the other hand between us
RPRians let us put together the essential mantra that might
lead us to the ulimate one liner.
So, if we have to say "RPR is not ethernet but uses the
ethernet physical layer" - we sound giberish and ashamed.
We also are touting that ethernet cannot do rings. This is a dangerous,
slipery slope because ethernet can be convincing that there is no need to do
shared rings.
Look, physical layers have been reused by many layer 2 protocols
but none have used that as the main enticement to lure people to
the new technology. The bottom line is that RPR needs to stand
on its own 2 feet. One of it is resiliency. The other one cannot be
because we need to wear shoes on both feet - viz: use ethernet or SONET
as a physical layer - it must for some other valid ratonale.
RPR is completly different in paradigm than ethernet. They both are packet
switched technologies. However, you cannot introduce RPR as "this is Darrel
and my other brother (ethernet) Darrel." If we do this we will reinforce
the dot3'ers perspective - that this just a few vendors who are stuck with
a propriatary protocol and trying to shoehorn it into a standard and masquerade
as ethernet in the press.
What we need is a simple message that captures the following:
1. RPR is a switching protocol for shared packet switched rings and not just a media
access protocol. Its more of a "network" access protocol.
2. RPR provides fine grain selective protection instead of doubling the bandwidth
So, this is not like throwing the big lever for switching tracks for the
entire train using spanning tree or like dumb BLSR in SONET.
3. RPR operates over currently available physical layers and hence has the
economic proposition in place to guarantee scaling the technology over time.
4. RPR has a more wholistic mechanism for managing network capacity and
associating it with multi-services
Now, lets get the marketing juices flowing......
RPR is all about Rings, Packets and Resiliency.
raj
-----Original Message-----
From: Dr. Gary Nelson [mailto:gnelson@xxxxxxxxxx]
Sent: Thursday, November 16, 2000 7:20 AM
To: Lauren.Schlicht@xxxxxxxxxxxx
Cc: RDLove; owner-stds-802-rprsg@xxxxxxxx; stds-802-rprsg@xxxxxxxx;
'Mike Takefman'; Yongbum Kim
Subject: Re: Avoiding Confusion in the Marketplace
It would be, I believe, a mistake to use the ethernet buzzword unless we
can justify on the grounds of
substance. People usually see through a hype fog fairly quickly and we
won't serve our own interests
by emulating other previous examples of overhyping.
It would be accurate to say that RPRs will support or transport or
tunnel "ethernet services including (but not limited to) VLANs".
Best regards,
Gary Nelson
Lauren.Schlicht@xxxxxxxxxxxx wrote:
>
> Not to mention "ethernet" is the buzz word in the market (whether we
> say NOT ethernet or ethernet rings) from a marketing PR point of view,
> we want to use the word if it fits.
>
>
> "RDLove" <rdlove@xxxxxxxxx>
Sent by: To: "Yongbum Kim"
owner-stds-802-rprsg@xxxxxxxx <ybkim@xxxxxxxxxxx>, "'Mike
Takefman'" <tak@xxxxxxxxx>,
11/15/00 05:32 PM <stds-802-rprsg@xxxxxxxx>
cc:
Subject: Re:
Avoiding Confusion in the
Marketplace
>
>
>
>
> I have a slightly different perspective on this. I certainly want to
> make
> sure the market is not confused by what RPR is, or that it be confused
> with
> Ethernet. However, I expect the word "Ethernet" will come up a lot in
>
> describing RPRs capabilities in words users understand. The reasons
> are
> twofold. We may be borrowing from the Ethernet physical layer, and we
> are
> likely to be using RPR rings to carry traffic that originates at
> Ethernet
> nodes. Because of the traffic we will carry, the ease of logiceal
> attachment of those nodes to RPR rings will be crucial and Ethernet
> compatibility will be a vital characteristic. But RPR will be more
> than
> Ethernet, and have capabilities that are fundamentally missing in
> Ethernet
> because of its architecture.
>
> I don't know if "not just an Ethernet Ring", or "More than just an
> Ethernet
> Ring" would be suitable one-liner descriptions of RPR. However, there
> is a
> great deal of truth in those one-liners, and I wouldn't eliminate them
> until
> a better candidate emerged.
>
> Best regards,
>
> Robert D. Love
> President, LAN Connect Consultants
> 7105 Leveret Circle Raleigh, NC 27615
> Phone: 919 848-6773 Mobile: 919 810-7816
> email: rdlove@xxxxxxxx Fax: 720 222-0900
> ----- Original Message -----
> From: Yongbum Kim <ybkim@xxxxxxxxxxx>
> To: 'Mike Takefman' <tak@xxxxxxxxx>; <stds-802-rprsg@xxxxxxxx>
> Sent: Wednesday, November 15, 2000 6:57 PM
> Subject: RE: Avoiding Confusion in the Marketplace
>
>
> >
> > Dear all,
> >
> > I understand the marketing power of the word "Ethernet". Let's
> > NOT use it anywhere. So far, all IEEE and ANSI Rings has been Token
>
> > based, and RPR is a ring without a token. But RPR MAC is Ring MAC.
> It
> > must discover and aware of its neighbors for forwarding decisions,
> access,
> > CoS control, etc. IEEE 802.3ae 10G is full-duplex only, essentially
> frame
> > format has been preserved (i.e. no CSMA/CD). RPR frame formats are
> different
> > from this. There is no Ethernet in RPR.
> >
> > Having said that, there are a lot of attractive characteristics of
> > Ethernet in this token-less Ring. It is truly distributed ring
> access,
> > plug and play, clear, simple (compared to 802.5, for example) and
> explicit
> > ring discovery MAC; full compatibility to 802.1's transparent
> bridging and
> > VLAN tagging; full compatibility to 802.3ad's port aggregation (if
> we
> > wish to entertain this); and the fact that RPR is optimized solution
> for
> > the problem (just like Ethernet was the optimal solution in its Coax
>
> days).
> >
> > RPR is a Ring Access MAC that is as easy to use as Ethernet, but not
>
> > Ethernet.
> >
> > regards,
> >
> > Yong.
> >
> > -----Original Message-----
> > From: Mike Takefman [SMTP:tak@xxxxxxxxx]
> > Sent: Wednesday, November 15, 2000 12:27 PM
> > To: stds-802-rprsg@xxxxxxxx
> > Subject: Re: Avoiding Confusion in the Marketplace
> >
> >
> > The idea of a tag line is a good one. Somewhat tongue in
> > cheek I suggest "RPR: Its NOT just Ethernet in a ring".
> >
> > To some extent I look to the Alliance to come up with a
> > good line since they are the marketing types.
> >
> > mike
> >
> > Luc Roy wrote:
> > >
> > > Mike,
> > >
> > > You have a good point. We should agree and adopt a one liner.
> It's not
> > > Ethernet but it is often viewed as Ethernet on a ring topology
> with
> > > Sonet/SDH performance and protection mechanisms. Simply RPR may
> not cut
> it.
> > > How about something simple like "RPR: Ethernet on a ring". I
> don't
> want to
> > > suggest that this is the right one-liner but something simple and
> > > descriptive is required, or, people will create their own general
> > > description which will confuse the market.
> > >
> > > Luc
> > >
> > > -----Original Message-----
> > > From: Mike Takefman [mailto:tak@xxxxxxxxx]
> > > Sent: Tuesday, November 14, 2000 7:01 PM
> > > To: stds-802-rprsg@xxxxxxxx
> > > Subject: Avoiding Confusion in the
> Marketplace
> > >
> > > People,
> > >
> > > For those of you who were not at the 802 plenary
> last
> > > week I want to inform you of an issue that we will
>
> > > have to watch diligently.
> > >
> > > Members of 802.3 expressed concern that the
> "market"
> > > will confuse 802.17 with 802.3 because we plan to
> > > re-use some of the physical layer work done in
> > > 802.3.
> > >
> > > I publicly stated to 802.3 that the intent of the
> > > RPRSG and by extension 802.17 (should we be
> approved)
> > > will be to make crystal clear to everyone that
> > > RPR != Ethernet even though we may share the same
> > > PHY layers.
> > >
> > > Let me be blunt, we do not want to have any sort
> of
> > > flap with 802.* over this issue. Inspite of
> concern
> > > over this issue they voted to allow our creation
> and
> > > I do not want to betray that trust.
> > >
> > > Under no circumstances should any company or
> individual
> > > (working in 802.17 or affiliating itself to
> 802.17)
> > > suggest that RPR == Ethernet. Any reference to
> Ethernet
> (be
> > > it
> > > 1GE or 10GE) should be that RPR will use the same
> > > physical layer as 1 or 10 GE.
> > >
> > > I would appreciate that people forward this
> message to
> > > their company's marketing people and ask that I am
>
> > > contacted if there is any confusion. I am not
> asking to
> > > become a clearinghouse for announcements, as there
>
> > > is a clear conflict with my role as a Cisco
> employee.
> > >
> > > I am asking that all of you work to insure we do
> > > not annoy our colleagues in other 802 groups.
> > > If any of you come across any public material that
> is
> > > questionable, please forward me a pointer to it
> > > ASAP.
> > >
> > > thank you for your attention,
> > >
> > > mike
>
>
--
Dr. Gary A. Nelson
Zynrgy Group Inc
20708 Deerpath Road
Barrington, IL 60010-3787
USA
+1.847.304.0000
+1.847.304.1929 fax
gnelson@xxxxxxxxxx