At 03:27 AM 4/2/2003 +0000, Priscilla Oppenheimer wrote: >I wonder if Cisco's MPLS class is just dated. It takes a long time to >develop and roll out a new class, especially if there's also a Cisco Press >book, exam, instructor materials, course binder, instructor training, beta >testing, etc.
More than likely, Cisco chose to teach what a broad range of their gear could do. L2vpn doesn't fit this category, though I would expect that they have better luck with RSVP. >In the early days of MPLS, was there more emphasis on LDP than on RSVP-TE? I find the two technologies not competitive actually. I am just now building a network that runs LDP on a large number of devices for ease of provisioning, yet rides a TE core that is signalled by RSVP-TE. To me, these are two tools. However, I agree with nrf that glossing over RSVP will leave a bit of a hole in one's knowledge. I again expect that Cisco may have had wider platform support for LDP than they did for RSVP, but I'd have to check that out as I know they were an early supporter of RSVP, but may not have offered it beyond their 7500/12000 product lines. >Were MPLS L3 VPNs around before L2 VPNs? RFC2547bis, or BGP/MPLS VPNs, was the first widely inter operable vpn technology that used MPLS in the forwarding plane. It is thus also the most mature of the many variants and again more widely support across the product line. L2vpn (ptp) is still pretty fresh, particularly in the Cisco camp. Very few platforms have a wide range of support for the many encapsulations defined by the various martini specs. (Luca Martini from L3 has taken the lead on the many L2 over MPLS encap standards as well as defined a signalling mechanism via LDP) I expect the standard course gear doesn't have enough support for these technologies to make labs feasible. I should note that the L2vpn (if you want to call it that and most marketing types do) I've been discussing (though briefly) are the point to point type (Virtual Private Wire Services -VPWS). Think frame relay with ethernet in the last mile and 802.1q tags for DLCIs. There are also a set of standards dealing with point to multipoint delivery, usually known as Virtual Private Lan Services that are attracting a bunch of a attention. These specs made the provider network look like a single broadcast domain. I'm not convinced that is a good thing (don't know many providers using LANE for what its worth), but it certainly seems exciting to marketing and IETF types. Anyway, I suppose my overall point is that I fully agree with nrf, that to the curricula is not entirely representative of the more interesting bits of MPLS, however I expect the underlying reason is lack of platform/sw support to enable effective classroom lecture on the subjects. Pete >Maybe it's just a matter of "course development latency." Thanks for your >insights. > >Priscilla > > >nrf wrote: > > > > ""Henry D."" wrote in message > > news:[EMAIL PROTECTED] > > > I don't mean to start any type of argument here, especially > > with someone > > > who obviously has more experience than I do. Yes, you've been > > > contributing to this study group many times. But also many > > times > > > your contributions are rather rethorical than practical and > > at the same > > > time you seem to draw attention to what your opinion is > > rather than to > > > give an educated and objective view backed by any type of > > real life > > > examples. > > > > First of all, given the subject matter (MPLS), it is most > > difficult to be > > giving out real-life examples. The fact is, MPLS is at this > > time not widely > > implemented, so therefore few examples abound. > > > > Second of all, it is essentially impossible for anybody to make > > a posting > > that is not necessarily colored with an opinion, particularly > > when they are > > discussing a subjective question. Questions like whether they > > should study > > MPLS or what they should do with their future are necessarily > > going to draw > > a wide range of opinions. If everybody is supposed to > > dogmatically answer > > 'yes' or 'no', then what's the point of even asking the > > question in the > > first place? The point is that subjective questions must > > necessarily elicit > > subjective answers. People are not robots. Everybody has to > > call it like > > they see it. You ask a subjective question, and people should > > be able to > > chime in with whatever they think. It's all about freedom of > > speech. > > > > Third of all, Cisconuts and I have taken the discussion > > offline, and while I > > don't want to speak for him, I would venture to say that he is > > quite happy > > with my responses. So if he's cool, then what exactly is your > > beef? > > > > Fourth of all, I resent the implication that my views are not > > educated. Be > > careful when you go around saying stuff like that. I seem to > > recall a story > > a few years ago how one particular guy harangued another guy > > about BGP, > > essentially saying that he knew nothing about how BGP really > > worked - only > > to find out later that the second guy was none other than a > > certain Tony Li, > > the father of BGP. Now, don't get me wrong, I'm not saying > > that I'm Li or > > anywhere close to him. What I'm saying is that you should > > watch your fire. > > > > >So yes, I'm saying that some times you don't quite stick > > > to the subject at hand. I don't see how your view on Cisco's > > curriculum > > > in re to MPLS can be taken seriously without you putting > > actual examples > > > of how you came to that conclusion. > > > > Ok, fine, then let's review the CCIP curricula vis-a-vis MPLS, > > and in > > particular, let's review what exactly they teach. I know for a > > fact that > > they teach primarily LDP and gloss over RSVP-TE. Do you think > > this is wise? > > There is no evidence in the industry of a consensus that LDP > > will > > automatically win out over RSVP-TE. If you have such evidence, > > I would like > > to see it. I doubt that LDP will ever win out simply because > > you can't do > > TE with LDP unless you go with CR-LDP which Cisco does not have > > any plans to > > support at this time. TE is one of the more important features > > available > > within MPLS. The point I'm making is that neglecting RSVP-TE > > within an MPLS > > exam seems rather dubious. > > > > Second, the last 2-3 modules of that class deal specifically > > with l3vpn's, > > with nary a mention of any l2vpn technology whatsoever. Again, > > why such an > > emphasis on L3 but no discussion of L2? Much of the most > > exciting work in > > MPLSCON is about l2vpn's. Don't get me wrong, L3 is good to > > know, but a > > good MPLS class would also get into a discussion of l2. > > > > The point I'm making is this. If all you do is follow the > > official Cisco > > MPLS class, you will get a warped view of how real-world MPLS > > is. LDP is > > not the ultimate no-brainer signalling path for constructing > > LSP's and MPLS > > can do far more than just L3VPN's. I'm not telling you not to > > follow > > Cisco's curricula. What I'm saying is that you should > > supplement it with > > other readings and experience. > > > > >Even if the knowledge required for > > > achieving > > > Cisco's recognition in re to MPLS was not as advanced as one > > would hope, > > > shouldn't we look at positives of the whole process ? > > > > Again, it's not a matter of being advanced as it has to do with > > emphasis. I > > think that the coursework emphasizes some of the > > not-so-important things and > > does not discuss some of the more important things. > > > > Also, I don't think it's my job to 'play nice'. If things are > > not good, > > then I think people should say that they're not good. Why > > engage in > > diplomatic euphemisms? Does it really do anybody any good to > > dress things > > up so that they look better than they really are? I'm not > > running a > > marketing campaign. > > > > > There are still things > > > to be learnt, and emphasising them rather than the weaknesses > > would be a > > > better idea. You won't become an expert just by passing the > > test or taking > > a > > > trainig > > > class, but at the same token, you can still learn a lot while > > achieving > > > those CCXX > > > goals. > > > > > > Anyway, I'm sure there will be a good response coming, so let > > me be done > > > with this subject. I had an early start today and I'm tired > > now. > > > > > > Good night ! > > > > > > > > > ""nrf"" wrote in message > > > news:[EMAIL PROTECTED] > > > > ""Henry D."" wrote in message > > > > news:[EMAIL PROTECTED] > > > > > Let me say up front, I don't have much experience in > > MPLS, I have > > > > > only played with it in the lab and not all that > > extensively either. > > > > > But CN is simply trying to get an idea of what to expect > > to go that > > > road. > > > > > > > > I believe that was precisely what I answered. > > > > > > > > > Is "nrf" saying not to advance in this field by studying > > Cisco's way > > of > > > > > emphasising MPLS ? > > > > > > > > What I said is that if you want to advance in that field, > > you will need > > > > substantially more than what Cisco wants you to know about > > it. Read my > > > post > > > > again. > > > > > > > > >You know, we all have our doubts, he's brave enough > > > > > to come to this group and ask questions. As far as > > L3VPN's, why not > > > > > concentrate > > > > > on that at least to start with. > > > > > > > > I never said not to learn L3VPN's. Read my post again. > > What I said is > > > that > > > > study of L3VPN's shouldn't be emphasized to the degree that > > Cisco seems > > to > > > > emphasize it. > > > > > > > > > It's still one reason to do the MPLS thing. > > > > > By just > > > > > doing that he'll need to touch on many aspects of MPLS > > anyway. He will > > > > still > > > > > use either LDP or RSVP, he still will use the LSP > > establishment, he > > > might > > > > as > > > > > well > > > > > learn the TE options available for establishment of those > > LSP's. He'll > > > > need > > > > > to learn > > > > > how to use the LSP's for pushing traffic over them. He'll > > learn what > > and > > > > how > > > > > the > > > > > labels get pushed/popped. Then why not study it that way. > > He's not > > > > advancing > > > > > his > > > > > MPLS skills, he might not have any yet. He's simply > > trying to see if > > he > > > > will > > > > > be able to utilize any of the skills he will have to > > learn to make it > > > > worth > > > > > it his while. > > > > > > > > No doubt all learning is good. Again, read my post again. > > I never said > > > > that he shouldn't learn it. What I said is that he > > shouldn't > > necessarily > > > > learn it "the Cisco way". > > > > > > > > > > > > > > Well, maybe someone else with more experience in MPLS > > arena and > > someone > > > > more > > > > > objective can give a better insight as to whether there > > is a demand > > for > > > > > these skills. > > > > > > > > Are you implying that I'm not objective - that I have some > > kind of > > agenda? > > > > > > > > > > > > > > > > > > > > > > > > ""nrf"" wrote in message > > > > > news:[EMAIL PROTECTED] > > > > > > ""Cisco Nuts"" wrote in message > > > > > > news:[EMAIL PROTECTED] > > > > > > > Hello group, How does one feel about a career in > > MPLS...I mean > > doing > > > > > MPLS > > > > > > > as part of your core job day in and out.....Is it > > worth it? Since > > > our > > > > > > > network does not use MPLS (maybe never will) inspite > > of being one > > of > > > > the > > > > > > > Big Four Tier 1 SP's.... > > > > > > > > > > > > Let me guess. Do you work for Sprint? > > > > > > > > > > > > >are there other SP's that use MPLS in their > > > > > > > backbone?? > > > > > > > > > > > > Yeah, there are some. > > > > > > > > > > > > >I have just given myself a month or so break from my > > CCIE Lab > > > > > > > Prep.(yeah!yeah! most would consider me stupid on > > this) to study > > > MPLS > > > > > > > for the CCIP and am thinking if I should pursue this > > subject just > > > > like > > > > > I > > > > > > > did for BGP.....know it inside out cold.....and maybe > > consider a > > new > > > > > > > career/job in MPLS (obviously along with BGP, MBGP, > > MCast etc...) > > > Does > > > > > > > anyone know of how MPLS is viewed out there? I > > mean, in terms of > > > > > > > implementation, popularity and last but not the least > > , $$$ ??? > > > > > ;->Which > > > > > > > of the Big SP's or Enterprise networks have > > implemented MPLS? Has > > it > > > > > been > > > > > > > worth the advantages that MPLS proposes??Thank > > you.Sincerely,CN > > > > > > > > > > > > The way I see it is this. MPLS is potentially powerful > > technology > > for > > > > it > > > > > > can be used as a lingua-franca among a carrier's > > network and > > transport > > > > > layer > > > > > > and also as a way to impose circuit-switching > > discipline upon IP and > > > > > > therefore offer circuit-switching services with a pure > > IP network. > > > > > > > > > > > > But MPLS is by no means a slam-dunk. Certain > > carriers, most > > notably > > > > > > Sprint, have elected not to go down the MPLS path > > because they > > believe > > > > the > > > > > > technology is immature (and they are correct) and also > > because they > > > > > believe > > > > > > that they can garner the benefits of MPLS by other > > means (also > > > correct). > > > > > > The point is that while MPLS offers great potential, it > > also > > presents > > > > > > problems, so implementing it is not a no-brainer. > > > > > > > > > > > > And furthermore, I don't particularly like the way that > > Cisco is > > > pushing > > > > > > MPLS, particularly in its cert program. In my opinion, > > I think > > > Cisco's > > > > > cert > > > > > > programs emphasize the least useful parts of MPLS while > > neglecting > > the > > > > > more > > > > > > useful parts. For example, I don't understand why > > Cisco pushes LDP > > > the > > > > > way > > > > > > it does, for LDP merely builds LSP's that correspond to > > the route > > > table, > > > > > but > > > > > > what's so useful about having LDP's that look like the > > route table? > > > It > > > > is > > > > > > far more useful to build LSP's that differ from the > > route table, but > > > the > > > > > > methods of doing that are not really covered very much > > (if at all) > > in > > > > the > > > > > > Cisco curricula. Also, I don't understand why Cisco > > places such an > > > > > emphasis > > > > > > on L3VPN's, as if L3VPNs were the only important > > service that MPLS > > > > > enables. > > > > > > L3VPN's are only one of the new services that you can > > enable, and in > > > my > > > > > > opinion, one of the less important ones. Far more > > important are the > > > > L2VPN > > > > > > capabilities and the ability to unify IP, ATM, and > > optical into a > > > single > > > > > > management plane. The point I'm making is that if > > you merely > > study > > > > MPLS > > > > > > according to the Cisco curricula, you really haven't > > learned much > > > about > > > > it > > > > > > that's actually useful. > > > > > > > > > > > > > > > > > > > > > > > > > > > > ------------------------------------------------------------------------ > > > > > > > > > > > > > > Add photos to your messages with MSN 8. Get 2 months > > FREE*. Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=66681&t=66609 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
