The networking industry loves good debate, gossip and scandal. Enterprise deals are full of rumours, innuendo, and the constant reshaping of vendors' images.
Unfortunately, the never-ending jousting between networking vendors often leaves the so-called hard questions unanswered. And in particular, the real issues worrying users can be buried on the field of battle.
Last week, the networking industry did its best to unearth these issues in a lively debate at an exhibition and conference in Sydney. It featured the managing directors of Australia's most prominent networking vendors.
Cisco's Gary Jackson, 3Com's Gerhard Rumpff, Cabletron's Ian Fewtrell and Nortel Networks' Steve Rust entered the jousting ring, with veteran IT journalist Beverley Head the adjudicator.
The action produced a mixed reaction from the audience. Some enjoyed what amounted to light entertainment, while others shook their heads, claiming the forum failed to address the real issues.
And in one sense the critics were right. In discussions with Network World, all but one of the vendors claimed they were not given an opportunity to answer each of the 'hard questions' the other three competitors prepared for them. So, as great believers in the 'fair go' ethic, we interviewed each vendor after the debate, armed with a leaked copy of the official (vendor anonymous) debate questions.
Q: Enterprises and service providers are increasingly asking for networks that can service both voice and data applications. How does Cabletron expect to succeed in these changing markets without broad experience and skills in telephony and IP routing?
Fewtrell: When we bought Digital, it was actually one of the first companies, if not the first company, to produce routers. We bought all that expertise. And secondly, we had very strong alliances with Ipsilon, which was the very first company to bring out IP switching. In addition to that, last year we bought Yago.
In terms of our experience [in voice], Nortel themselves use Spectrum [Cabletron's network management product] to manage their network.
Q: How is Cabletron planning to rationalise its disparate and overlapping product and technology offerings [Digital and Cabletron's], given the enormous cost associated with support and R&D, whilst maintaining relevance in the ever more competitive enterprise market?
Fewtrell: My answer is: Why rationalise? We have just grown by 287 per cent and you don't rationalise something that's growing for you.
Q: Jack Welsh, CEO of GE, manages by the rule: number one or two in any chosen market or get out. He successfully transformed GE using this tenet. If you are not number one or two in your chosen (current) market segment now, in which markets will you achieve this status in the next three years?
Fewtrell: GE has just signed a reseller agreement with Cabletron, so it obviously recognises success. And if it is adopting the same view, I must be one or two. [However] the real issue is: Are you [number] one or two in the right market?
What about the growth markets? We are number one in Layer 3 and Layer 4 switching [according to IT analysts Dell'Oro].
Basically, the markets that we selected are the gigabit market, Layer 3 and Layer 4 switching and enterprise management. And in two of those we are already recognised as in the top two, as indicated by top surveys.
Q: How does 3Com intend to compete in the service provider space without the ability to offer a more complete portfolio of access, multiservice networking, high-performance switched routing, and high-capacity optical transmission products?
Rumpff: Pretty simple answer. Everything, bar the high-capacity optical transmission products, we've got. [In the access market] we've got PathBuilder, we've got Total Control. At the top end we've got CoreBuilder, so I don't know where that question came from. As for optical transmission technology, that is the domain where very few of the traditional suppliers are in anyway. But we have equipment that will help organisations to double their fibre capacity quite easily. So the whole question as such is not correct.
Q: With 55 per cent of your revenue coming from embedded NICs (a market rapidly being consumed by the PC manufacturers) and combined with 3Com's continuing lack of success in the enterprise arena, is your strategy now to cut your losses and prove the pundits right that 3Com is a low-end player that now must attempt to concentrate on the modem market by your recent advertising campaigns?
Rumpff: On a worldwide basis, our total revenue from NICs is about 25 per cent. In Asia/Pacific, our revenue out of NICs is close to 28 per cent. In Australia, our total revenue for NICs is 29 per cent. This [question] is totally wrong.
Well over 35 per cent of our revenue comes out of the enterprise area. One of the reasons why we have made a lot of noise in the marketplace about some of the lower end products is purely because we see a tremendous opportunity to take market share and leadership. The enterprise game is a relationship game where you work with large enterprise accounts. That is not a marketing exercise.
Q: The world is going through a serious IT skills shortage. What demonstrable action is your company taking to assist the industry address this issue?
Rumpff: This is an issue the whole industry is facing. One of the key issues we need to address to overcome this situation . . . is really to make that whole area of networking and IT usage much easier.
What we are driving into the marketplace from a networking point of view is zero administration. We need to make it plug and play, and we as a company are working on it today.
Q: Cisco has tried to position IOS as a true operating system. However, it has none of the characteristics or capabilities commonly found across operating systems in the industry. Commercial operating systems like Windows, for example, are open, have well-defined APIs, offer development tool kits, and support a huge ISV [independent software vendor] community. If IOS is a true operating system, why doesn't Cisco open it up to the entire networking industry? Or are they determined to maintain proprietary systems which lock in customers?
Jackson: The first thing is, we actually licensed IOS and have been prepared to make it available to anyone under licence. In fact, Compaq was one of those to purchase a licence. We have always supported open connectivity to IOS; we developed a tool kit you can get free of charge. The main reason it's not 'taking up like Windows' is if you look at desktop standards and open operating systems, they tend to get driven by users who want to write around them. When you are talking about a plumbing operating system, the only people who are really interested in it are those who are offering alternative plumbing environments.
Q: In the new world of voice and data convergence, strategic alliances and partnerships between data and voice networking vendors are vital to deliver the needed solutions. How will Cisco, with its 'go-it-alone' strategy, accelerate its R&D without access to voice technology (through to PABX) and the high risk of infringing existing patents?
Jackson: I think the real basis of [this question] is an incorrect positioning. It says we are not going to get access to voice technology. Well, that's just plain incorrect. We have access to voice technology, it's just that the voice technology we are looking to develop and incorporate is IP-based. It is not the old 64Kbit/sec cross connect-type technology, that frankly we are not interested in.
We currently do have alliance proposals with several of our competitors when that's what we think our customers are demanding.
In terms of the go forward strategy on voice technology, I fail to see how we are going to be breaching anybody's patents when we are looking at an IP-based voice and data integrated world.
Q: Marketed as homogeneous yet so obviously disparate, the Cisco product range has many varieties of IOS, and at least four distinct management products (a consequence of growth by acquisition), and the total cost of ownership to the end user can only continue to rise. Is the answer yet more discounting?
Jackson: The first point is, IOS is a modularised operating system. The fact that it covers all our products, you could say it is disparate.
However you could also say it is a single operating system quite intentionally from Cisco to cover our full range of products to have one integrated operating system.
So disparate to me is an emotional word, rather than a logical one.
There is no doubt that Cisco does need to provide an overall, what we call full flow-through provisioning of network management from the top end to the bottom end.
Regarding discounting, our cost of ownership over a five-year period is highly competitive, and in addition we have got (better) support.
Q: With the increasing importance of the SME (small to medium enterprise) market and the extensive channel needed to cover this market, what is Bay-Nortel's 'go-to-market' strategy in Australia, given the ever-decreasing channel presence of Bay (JNA, Digital) and Nortel's 'direct' policy?
Rust: We do have products that address the SOHO/SME market. We've got two: the BayStack, and a new product range called NetGear for the SOHO market we introduced about 18 months ago. It's already running at a multimillion dollar run rate, and it's all going through the channel.
Q: With Lucent acquiring part of your channel (JNA); loss of identity and turmoil due to the recent takeover; Nortel passing over Bay products in preference for Cisco's to provide a solution to Powertel; and negative revenue growth, is there a future for Bay in Australia?
Rust: I see the future world around applications that use voice data and video and the companies that are positioning to enter that market place strongly are companies like Nortel Networks. Our Bay Networks line of business grew by 30 per cent last year and our channel continues to grow very strongly.
As for Powertel, Nortel Networks put forward a Nortel solution and it didn't include any Cisco product at all.
Q: What is your company's present and projected future capabilities in the area of data, voice and video using IP as the transmission protocol?
Rust: We see the world on IP, and Bay Networks has already made a significant investment here with our Symposium Call Centre. We believe carrier-grade applications are going to be the applications of the future. I also believe we are well positioned to get across voice, video and data.
Users want applications that are as easy to use as a telephone, and that isn't there yet in the data world.