From: Colin McNamara (colin@2cups.com)
Date: Wed Aug 22 2007 - 21:30:29 ART
John Chambers is keynoting VMware World in a couple weeks. My guess is
that your questions will be asked then. Sadly, I don't have a hard
answer for you know.
You are correct, that the vswitches that are currently in vmware are
just bridges, that are mapped to physical or logical sub interfaces on
the ESX server. Normally what you do is configure your ESX servers with
dot1q interfaces, so you can trunk up into an aggregation layer. It does
however leave much to be desired. The scariest part is, with
Virtualization taking such a strong hold, we really have minimal
visibility into the new network edge. My guess is that Cisco will
provide the features we need, such as deterministic configration and
behavior, spanning, layer 3 services, etc.
I think this also fits into the larger vframe (data center
virtulization) solution from Cisco. By putting intelligent devices back
at the edge it enables far more advanced solutions to be deployed
without multiplying the development effort.
--Colin
On Wed, 2007-08-22 at 13:22 -0700, Scott Vermillion wrote:
> Very interesting Colin. I worked a project last year where I did the
> physical network part of a project and another sub did the virtual
> machine stuff. Thus, my exposure was more indirect than anything (I had
> my own hands full, couldn't really ghost over those guys as much as I'd
> have liked). What I recall was that they had a virtual switch for
> inter-VM connectivity w/in the physical server. It was a basic "frame"
> mover w/out much in the way of fancy capabilities. Just a GUI with the
> capability to drag connections between VM instances.
>
> So Cisco sees a role for themselves in this arena. I guess I can see
> where that might be the case. But I'm not clear whether or not this will
> have direct value for CCIE candidates? I suppose if their virtual
> switches are made with the intention of connecting out to the physical
> network (and now that I think about it, why wouldn't they?), then they
> would need to support many, if not most, of the major capabilities that
> real switches do. Very interesting, indeed. I wonder when this will hit
> the street and on what kind of platforms this will be able to be ported
> to? Will you be able to run a stand-alone VS (virtual switch?) or will
> such a thing only be bundled with other licensing? Etc?
>
> -------- Original Message --------
> Subject: Re: Virtual CCIE's? - 3550/3560
> From: Colin McNamara <colin@2cups.com>
> Date: Wed, August 22, 2007 1:56 pm
> To: ccielab@groupstudy.com
>
> I don't know if anyone has seen the press-releases but Cisco is
> developing a virtual switch for VMware. There are no hard stats on
> features yet, but what has been released hints to at minimum layer 3
> switching in the virtual switch being made for ESX. I would guess
> that
> if you fast forward to next year, many peoples labs will consist of
> Dynamips (probably running on a vmware image) integrated with these
> virtual switches.
>
> --
> Colin McNamara
> (858)208-8105
> CCIE #18233
> "The difficult we do immediately, the impossible just takes a little
> longer"
>
> On Tue, 2007-08-21 at 19:42 +0100, Gary Duncanson wrote:
> > Makes sense.
> >
> > Thanks David.
> >
> > Gary
> > ----- Original Message -----
> > From: "David Prall" <dcp@dcptech.com>
> > To: "'Scott Vermillion'" <scott_ccie_list@it-ag.com>;
> > <ccielab@groupstudy.com>
> > Sent: Tuesday, August 21, 2007 5:58 PM
> > Subject: RE: Virtual CCIE's? - 3550/3560
> >
> >
> > > SIMICS is hardware emulation. It actually emulates the processors
> and
> > > asics.
> > > Putting together a switch would be rather difficult unless you
> worked for
> > > Cisco since the ASIC's are proprietary and the processor to asic
> linkage
> > > is
> > > part of the design. Also a license for Simics is rather
> expensive.
> > > Packages
> > > like this are typically used in the design phase, since an ASIC
> can be
> > > simulated rather easily, where as having it produced for
> debugging
> > > purposes
> > > can take months. So the software emulation of hardware is well
> worthwhile
> > > during the design phase.
> > >
> > > --
> > > http://dcp.dcptech.com
> > >
> > >
> > >> -----Original Message-----
> > >> From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On
> > >> Behalf Of Scott Vermillion
> > >> Sent: Tuesday, August 21, 2007 12:28 PM
> > >> To: ccielab@groupstudy.com
> > >> Subject: RE: Virtual CCIE's? - 3550/3560
> > >>
> > >> The good news is that they appear to offer an "academic" license
> for
> > >> individual use (approval time ~1 week). The bad news is that
> when I
> > >> google "virtutech cisco switch," I come up with precious
> > >> little. Here is
> > >> a link to a presentation where Cisco is mentioned in passing:
> > >>
> > >> http://www.deserec.eu/files/first_workshop/pdf/DESEREC_IABG_Si
> > >> mics_Workshop_2006.pdf
> > >>
> > >> It may indeed be possible to emulate a Cisco switch, but it
> > >> appears as if
> > >> there exists no "Dynagen equivalent" for we mere mortals. And no
> real
> > >> history to indicate CPU/memory requirements (could you
> > >> emulate one switch
> > >> with a given hardware configuration or could you emulate
> > >> 12?). Certainly
> > >> there doesn't appear to be any "body of knowledge" out there.
> > >> Also, not
> > >> sure how to take it, but that presentation makes mention of only
> being
> > >> able to bridge "TCP and UDP" from the simulated environment
> > >> to a physical
> > >> NIC (so no control plane traffic directly over IP? etc?). Thus,
> I'll
> > >> keep my ear to the ground on this one, but I still plan to
> > >> spend the next
> > >> several months tackling the CCIE R&S lab vs. being one of the
> first to
> > >> figure SIMICS and Cisco switches out...
> > >>
> > >> -------- Original Message --------
> > >> Subject: RE: Virtual CCIE's? - 3550/3560
> > >> From: "Lamine BOUAFIA" <b_lamine@yahoo.fr>
> > >> Date: Tue, August 21, 2007 6:48 am
> > >> To: <ccielab@groupstudy.com>
> > >>
> > >> Anyone who know mode about
> > >> http://www.virtutech.com/products/ and how
> > >> to
> > >> simulate Cisco 3550/3560??
> > >>
> > >> Regards,
> > >>
> > >> -----Message d'origine-----
> > >> De : nobody@groupstudy.com [mailto:nobody@groupstudy.com] De la
> part
> > >> de
> > >> Brian Dennis
> > >> Envoyi : mardi 21 ao{t 2007 06:42
> > >> @ : Scott Vermillion; Cisco certification
> > >> Objet : [Bulk] RE: Virtual CCIE's?
> > >>
> > >> When teaching a class at Cisco recently one of the students
> > >> said that
> > >> the
> > >> 3550 or 3560 can be "virtualized" using software from this
> company:
> > >>
> > >> http://www.virtutech.com/products/
> > >> https://www.simics.net/
> > >>
> > >> Personally I never looked into it but someone here may be
> familiar
> > >> with
> > >> the software and it's capabilities.
> > >>
> > >> Brian Dennis, CCIE4 #2210 (R&S/ISP-Dial/Security/SP)
> > >> bdennis@internetworkexpert.com
> > >>
> > >> Internetwork Expert, Inc.
> > >> http://www.InternetworkExpert.com
> > >> Toll Free: 877-224-8987
> > >> Direct: 775-745-6404 (Outside the US and Canada)
> > >>
> > >> >----- Original Message -----
> > >> Subject: RE: Virtual CCIE's?
> > >> Date: Mon, August 20, 2007 21:30
> > >> From: "Scott Vermillion" <scott_ccie_list@it-ag.com>
> > >>
> > >> > Unfortunately, the more recent posts I've seen from him would
> > >> indicated
> > >> > that, due largely to ASICs that cannot easily be
> > >> reverse-engineered, a
> > >> > switch equivalent or extension is all but dead as a concept.
> This
> > >> is why
> > >> > I chose to go ahead and purchase some 3560-8PCs and just
> > >> move on...
> > >> >
> > >> > -------- Original Message --------
> > >> > Subject: Re: Virtual CCIE's?
> > >> > From: "darth router" <darklordrouter@gmail.com>
> > >> > Date: Mon, August 20, 2007 9:18 pm
> > >> > To: ISolveSystems <support@isolvesystems.com>
> > >> > Cc: "Scott Vermillion" <scott_ccie_list@it-ag.com>, "Cisco
> > >> > certification" <ccielab@groupstudy.com>
> > >> >
> > >> > That depends on the devs, and if they can continue to support
> > >> future
> > >> > platforms, whether it is possible or feasible. No switches
> yet,
> > >> maybe
> > >> > never, but who knows, chris has some posts where he wanted
> access
> > >> to
> > >> > 6500 switches. Might be workin on it :P I hope so. Here
> > >> is the site
> > >> > with the history.
> > >> >
> > >> > http://www.ipflow.utc.fr/index.php/Cisco_7200_Simulator
> > >> >
> > >> > On 8/21/07, ISolveSystems <support@isolvesystems.com> wrote:
> > >> >
> > >> > You said that Dynamips is in its infancy. I am curious to know
> > >> > how old is
> > >> > Dynamips? How do you see the continuing development of
> Dynamips
> > >> > to support
> > >> > future IOS development?
> > >> >
> > >> > Regards,
> > >> >
> > >> > On 8/20/07, Scott Vermillion < scott_ccie_list@it-ag.com>
> wrote:
> > >> > >
> > >> > > Gregory,
> > >> > >
> > >> > > Were a person to acquire the CCIE cert w/ nothing more than
> > >> > time on
> > >> > > Dynamips, that person may indeed face a few embarrassing
> > >> > moments
> > >> > > early on. So it should be outlawed and shunned by the
> > >> > CCIE-seeking
> > >> > > community at large? I'd wager that 90% or better using
> > >> > Dynamips for
> > >> > > CCIE study have sufficient experience w/ real HW to know
> > >> > which side
> > >> > > the power switch is on.
> > >> > & gt;
> > >> > > Do you think you'll normally have physical access to all of
> > >> > the
> > >> > > routers on which you are expected to perform your work? If
> > >> > you
> > >> > > require that to do your job, are you effective?
> > >> > >
> > >> > > Recall that this is real IOS -- not some training
> > >> > simulator. So the
> > >> > > chassis, the power supply, the interface cards, etc. are
> > >> > lacking in
> > >> > > your CCIE lab, where you are intensely focused on subtle
> > >> > protocol
> > >> > > interplay. So? Can you even touch the HW during the CCIE
> > >> > practical
> > >> > > exam?
> > >> > >
> > >> > > I have posted this as recently as a few hours ago on the
> > >> > professional
> > >> > > board:
> > >> > >
> > >> > > "I have now been dealing with Dynamips since (roughly) Dec
> of
> > >> > last
> > >> > > year. I generally trust it (yes, bugs now and then and even
> > >> > the
> > >> > > occasional crash, but can IOS itself claim otherwise?) and
> > >> > actually
> > >> > > prefer it over real hardware for a couple of reasons:
> > >> > >
> > >> > > 1. Recabling a HW lab takes longer than launching a new .net
> > >> > file. I
> > >> > > keep every .net file I've ever created, along with all of
> the
> > >> > router
> > >> > > NVRAM files, etc. Modifying one lab for another purpose gets
> > >> > easier
> > >> > > and easier as your collection grows. Take good notes on why
> > >> > you
> > >> > > created a given environment and your results. This becomes a
> > >> > highly
> > >> > > valuable reference resource when you've grown a little fuzzy
> > >> > on some
> > >> > > details of a prior battle.
> > >> > >
> > >> > > 2. The capture function of Dynagen is killer. Definitely
> > >> > beats debug
> > >> > > output when you're really trying to understand what's going
> > >> > on under
> > >> > > the hood. If you're proficient w/ Wireshark or any other PA
> > >> > that can
> > >> > > open a .cap file, this is a "ki ller app" for sur e."
> > >> > >
> > >> > > Dynamips is nothing short of revolutionary for
> understanding,
> > >> > > troubleshooting, prototyping, and generally poking and
> > >> > prodding
> > >> > > network-related protocols. It has some limitations, yet in
> > >> > its
> > >> > > infancy. Know those and live with them. And then leverage a
> > >> > tool
> > >> > > that can only be bested by some seriously deep pockets...
> > >> > >
> > >> > > Regards,
> > >> > >
> > >> > > Scott
> > >> > >
> > >> > > -------- Original MessageGr --------
> > >> > > Subject: Virtual CCIE's?
> > >> > > From: "Gregory Gombas" < ggombas@gmail.com>
> > >> > > Date: Mon, August 20, 2007 6:19 pm
> > >> > > To: "Cisco certification" <ccielab@groupstudy.com>
> > >> > >
> > >> > > Guys,
> > >> > >
> > >> > > I was checking out Dynamips and its pretty cool and all,
> > >> > but it
> > >> > > does
> > >> > > worry me a little bit....
> > >> > >
> > >> > > How will employers view the CCIE certification after
> > >> > they've been
> > >> > > burned by hiring a CCIE who has never touched a real router
> > >> > in
> > >> > > their
> > >> > > life?
> > >> > >
> > >> > > Do you like the idea of a pilot flying your plane whose
> > >> > only
> > >> > > training
> > >> > > was with a virtual flight simulator?
> > >> > >
> > >> > > ; I remember the days when the MCSE was a hot cert until an
> > >> > army of
> > >> > > paper CCIE's hit the job market.
> > >> > >
> > >> > > Maybe they won't call it a paper CCIE, maybe they'll coin a
> > >> > new
> > >> > > term
> > >> > > like virtual CCIE.
> > >> > >
> > >> > > Just food for thought...
> > >> > >
> > >> > >
> > >> > >
> > >> >
> > >>
> > >> ______________________________________________________________
> > >> _________
> > >> > > Subscription information may be found at:
> > >> > > http://www.groupstudy.com/list/CCIELab.html
> > >> > >
> > >> > >
> > >> >
> > >>
> > >> ______________________________________________________________
> > >> _________
> > >> > > Subscription information may be found at:
> > >> > > http://www.groupstudy.com/list/CCIELab.html
> > >> >
> > >> >
> > >>
> > >> ______________________________________________________________
> > >> _________
> > >> > Subscription information may be found at:
> > >> > http://www.groupstudy.com/list/CCIELab.html
> > >> >
> > >> >
> > >>
> > >> ______________________________________________________________
> > >> _________
> > >> > Subscription information may be found at:
> > >> > http://www.groupstudy.com/list/CCIELab.html
> > >>
> > >>
> > >> ______________________________________________________________
> > >> _________
> > >> Subscription information may be found at:
> > >> http://www.groupstudy.com/list/CCIELab.html
> > >>
> > >>
> > >> ______________________________________________________________
> > >> _________
> > >> Subscription information may be found at:
> > >> http://www.groupstudy.com/list/CCIELab.html
> > >>
> > >> ______________________________________________________________
> > >> _________
> > >> Subscription information may be found at:
> > >> http://www.groupstudy.com/list/CCIELab.html
> > >
> > >
> _______________________________________________________________________
> > > Subscription information may be found at:
> > > http://www.groupstudy.com/list/CCIELab.html
> >
> >
> _______________________________________________________________________
> > Subscription information may be found at:
> > http://www.groupstudy.com/list/CCIELab.html
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
This archive was generated by hypermail 2.1.4 : Sat Sep 01 2007 - 11:32:12 ART