Monthly Archives: July 2010

Fear of a Converged Data Center

In a relatively short piece today, Michael Vizard has managed to cover a lot of ground. He deserves plaudits for his concision.

Quoting Ashish Nadkarni, a practice lead for Glasshouse Technologies, Vizard’s salient point is that while vendors, notably Cisco and HP, are pushing data-center convergence with fiery ardor, enterprises have not responded with reciprocal fervor.

Resistance is Manifold

The resistance to data-center convergence is manifold. CFOs are wary of anything resembling forklift upgrades accompanied by substantial capital outlays. Meanwhile, CTOs and CIOs are leery of stumbling into vendors’ trapping pits, drawn by the promise of long-term cost savings into a dungeon of proprietary servitude.

Last, and definitely not least, there is cultural and political resistance to sweeping change within IT departments. This makes perfect sense. Any student of history will know that revolutions displace and supplant power structures. The status quo gets pushed aside.

If we think about data-center convergence, we find that many potential enterprise-IT interests are threatened by its advance. As Vizard has mentioned previously, IT departments long have had their specialists. They are staffed by high priests of servers, viscounts of storage, lords of networking, and a smattering of application wizards.

Kumbaya Falls on Deaf Ears

By its very nature, data-center convergence entails that all these domain masters work in concert rather than in isolation. That scenario has theoretical appeal, and many salutary benefits could result from such IT kumbaya and common cause.

However, human beings — particularly in a realm where their positions are subject to offshoring and where job security has faded into a bitter, mocking memory– can be forgiven for eschewing collective idealism in favor of realpolitik calculations of personal survival. In their minds, questions abound.

If the data center is converged, what happens to the specialists? Who benefits, who wins and loses, who emerges from the fray with a prosperous career path and who becomes a dead man walking? These are uncomfortable questions, I know. But you can be sure many people are asking them, if only to themselves.

Answers Needed

An integrated, unified data center, with across-the-board automation and single-console manageability, has its charms — some of which are undeniable — but not necessarily to the specialists who inhabit today’s enterprise data center.

Cloud computing, whether of the private or public variety, faces many of the same issues, though the public option addresses the CFO’s concern regarding capital expenditures. Then again, cloud computing is challenged by the same cultural and political issues discussed above, and by other inhibitors, such as nagging questions about security and compliance.

I know these issues have been discussed before, here and elsewhere, such as by Lori MacVittie at F5’s DevCentral. Vendors, especially executives ensconced in boardrooms eating catered lunches, tend to overlook these considerations. Their salespeople, though, need cogent answers — and they had better be the right ones.

Strategic Considerations of Juniper’s OEM Deals

As an old business-development hand, I am curious as to how Juniper’s increasingly significant OEM relationships with IBM and Dell might factor into strategic considerations.

In my experience, OEM relationships are balancing acts that never remain at rest. A permanent equilibrium is impossible to attain. They wax or wane, but they’re always subject to dynamic tension and latent volatility.

Such, I’m sure, is the case with Juniper’s OEM relationships with IBM and Dell. The latter relationship seems to be getting closer by the minute, as the latest blog post at Juniper’s The Network Ahead attests.

The OEM Dynamic

To understand what I’m getting at, one first has to consider why companies decide to OEM products rather than build them on their own or buy them through an acquisition. Typically, it’s because a company has a gap in its product portfolio that prevents it from capitalizing on a perceived market opportunity within its installed base of customers. That opportunity usually falls into an area where the company believes it has a mandate to play.

Still, the company is tentative, uncertain as to how much it’s wiling to put into the venture. It might not have the internal resources or institutional expertise to build the product on its own, or it might be reluctant to make the necessary investments to do so.

Similarly, it might not see enough core strategic value to warrant making an acquisition to procure products, technologies, and skill sets in question. It’s basically at a point where it feels it has a mandate to enter a market, and it believes it can do well selling into that market, but it’s unable or unwilling to build the product on its own and it’s not ready to make an acquisition.

That is when an OEM relationship comes into play. It explains how and why IBM and Dell have come to rebrand and resell significant swathes of Juniper’s product portfolio.

Status Quo Not Permanent

Those relationships can go one of two ways, but a permanent holding pattern probably is not in the cards. At some point, Dell and IBM, separately, might decide to stop carrying Juniper’s gear, for one of various possible reasons, or will attempt to acquire Juniper or another company that does some or all of what Juniper does.

Meanwhile, IBM and Dell, because they’ve each invested increasingly in a relationship with Juniper, will be concerned about Juniper’s existential status. What I mean is that, as IBM or Dell becomes more confident and proficient in the sale of networking gear, it will become more concerned about losing Juniper as an OEM partner. How would that happen? Through an acquisition of Juniper by another vendor, of course.

I’m sure executives at Dell and IBM tried (perhaps successfully) to insert restrictive clauses into their OEM deals with Juniper, attempting to account for any and all scenarios in which Juniper would be acquired by another party or otherwise cease to exist. Those contractual clauses and provisions can take many forms, but they’re all about mitigating risk. The objective is to avoid being jilted and spurned, left bereft at the altar with a gaping hole in your product portfolio and disaffected customers wondering how you’ll address their needs.

Changing Value Equation

Typically, such contractual clauses ensure that maintenance, support, and other residual services will continue in the event of the sale or dissolution of the supplier company; but there are other considerations, too.

For example, right of first refusal, in the event of a an acquisition bid from a third party, and acquisition veto clauses are two recourses that receiving OEM companies often pursue when negotiating deals. The latter usually is resisted vehemently and summarily rejected by the company providing the OEM products in question, but sometimes — if the deal is big enough and the circumstances warrant it — some consideration of the former (right of first refusal) is included in the deal.

So far, Juniper has done well in developing, managing, and maintaining its OEM relationships with IBM and Dell. The nature of those relationships is changing, though, as Juniper invests its Junos-based “3-2-1″ network architecture with increasing amounts of dynamic intelligence and automation. To the extent that Juniper is successful in impressing the value of its intelligent network infrastructure on IBM and Dell customers, IBM and Dell necessarily will take notice.

As the value equations change, so will the relationships.

Microsoft Leads Analysts Astray

Microsoft today will host its annual meeting with market analysts. The company will bring the visitors up to speed on strategic initiatives, discuss salient market and technology trends, spotlight key products and solutions, and perhaps reset or gently massage market expectations for the year ahead.

As I read what analysts had on their minds as they prepared for the gathering in Redmond, I lost hope that Microsoft finally would muster the courage to look itself the mirror and acknowledge the earnest business-solution purveyor that stares back at it. I was tempted to say that the analysts are part of Microsoft’s problem — that they’re focused on the wrong things, that they don’t understand the essence of Microsoft, that they don’t appreciate the company’s inherent strengths and weaknesses — but, you know, that just wouldn’t have been fair, much less right.

Analysts take their cues from the companies they follow. If the market watchers monitoring Microsoft are stumbling down a blind alley, that’s because Microsoft led them there, perhaps even setting the wrong GPS coordinates on a doomed Windows Mobile application.

It follows, then, that if the guests at Microsoft today are focused on the wrong things — if they’re looking for answers and guidance on markets where Microsoft shouldn’t be playing, where it should scale back its efforts and investments, or where it needs to rethink its strategy — the fault is entirely Microsoft’s. The analysts are preoccupied with Microsoft’s consumer-facing product roadmaps, revenue projections, margins, and earnings (or lack thereof) because that is where Microsoft has focused their attention.

Rather than pointing at its potential to expand its presence and to achieve further growth in its core business markets — SMBs and large enterprises, and where and how those constituencies will consume application and computing services in future — Microsoft perversely has chosen to showcase its embarrassments and warts. It’s not a pretty sight, as the Kin fiasco demonstrates.

Meanwhile, if Microsoft would only listen, its customers — even its closest partners — are trying to set it straight. Yesterday, for example, HP confirmed that it would pursue a dual-tablet strategy, providing a Windows 7-based tablet for business customers and a webOS-based tablet for consumers. HP knows where Microsoft is strong and where it’s not so strong.

Microsoft might get the message one of these days. I’m just not expecting the epiphany to arrive today.

Avaya’s Kennedy Sends Cautious Signals on Post-Nortel Business

Reading between the lines of Avaya CEO Kevin Kennedy’s recent interview with Network World, I have the strong suspicion that revenues from Nortel’s installed base of VoIP and unified communications (UC) customers are not ramping as robustly as Avaya had hoped they would.

I get that impression as much from what Kennedy doesn’t say as from what he says. He’s bold and brash when talking about combined R&D efforts and product roadmaps, but he’s reserved when discussing revenue targets and near-term sales. He doesn’t say the Avaya-Nortel combination has been a commercial disappointment, but he’s not boasting of its conquests, either.

A few market analysts are noticing that Avaya’s acquisition of the Nortel enterprise business hasn’t resulted in market-share hegemony for the merged company. These market watchers seem surprised that Avaya didn’t take the Nortel customer base by storm and leave Cisco in its rearview mirror, choking on dust and fumes.

But that failure to reconcile with reality is at least as much the analysts’ fault as it is Avaya’s. Earlier in this saga, I noted that a Nortel-fortified Avaya would be fortunate to maintain any market-share edge over Cisco. It seemed an obvious conclusion to reach.

Unfortunately, though, when unwary market analysts examine a post-acquisition scenario, they will add the market share of the two companies involved, then assume the merged entity will maintain or extend its combined market share. For many reasons, however, that rarely — if ever — happens.

In the case of Avaya’s acquisition of Norte’s enterprise business, several complicating factors suggested that the merger, from a market-share perspective, would result in less than the sum of its parts.

First, there was the product overlap, which was not insignificant. Second,  there were channel-management issues, which also were considerable. (Some Nortel partners were concerned about having to deal with Avaya.) Third, Nortel’s enterprise business had been in distress for some time, and it was suffering market-share erosion before and after Avaya took control. Fourth, even among Nortel customers still in the fold, some eventually will choose options other than those presented by Avaya.

I think Avaya anticipated most (if not all) of these challenges. Just after the acquisition closed, for example, Kennedy sought to temper post-merger expectations. He cited external factors, such as the weak economy, as well as the usual post-merger integration challenges. His tone was one of cautious optimism rather than of unchecked exuberance. He knew it wouldn’t be easy, with or without Nortel’s enterprise business.

He’s staying on message, probably for good reason.

Before Foxconn, Huawei Had Its Own Suicides

Long before the rash of deaths at the Foxconn Technology Group’s manufacturing facilities in China, another company fought to stem a wave of suicides at its Chinese operations.

That company was Huawei Technologies, and its problem with suicidal employees was covered in the media, though not as extensively as were the unfortunate events at Foxconn, part of the Hon Hai Precision Industry Co., Ltd.

What partly accounts for the difference in degree of coverage, I think, is Foxconn’s connection to Apple. As we all know, Foxconn manufactures Apple’s iPhones and iPads as well as computing devices for a number of other vendors, including Dell. Everything Apple touches is high profile, so it’s no wonder that the Western media gravitated to the Foxonn suicides once  Apple was discovered among Foxconn’s brand-name customers.

Another factor, though, might be the intense secrecy that surrounds Huawei. It’s a privately held company, shrouded in mystery, run by CEO Ren Zhengfei, who emerged from the People Liberation Army (PLA), is a member in good standing of the Communist Party of China, and is said to retain close ties to China’s defense and intelligence elite.

Still, the suicides at Huawei are a matter of public record. They began ramping in the year 2000 and continued well into the decade, seemingly coming to an end — or something like one — by 2008. At their peak, they were bad enough that Ren Zengfei wrote the following to another member of the Communist Party:

“At Huawei, employees are continuously committing suicide or self-mutilation. There is also a worrying increase in the number of employees who are suffering from depression and anxiety. What can we do to help our employees have a more positive and open attitude towards life? I have thought about it over and over again, but I have been unable to come up with a solution.”

This is not exactly the sort of pitch a human-resources executive wants to feature in employee-recruitment campaigns. Nonetheless, it demonstrates that Ren recognized the problem and was thinking hard about whether his company’s “wolf culture” and “mattress culture” were sustainable models on which to build a business that could scale and compete successfully against the world’s leading telecommunications-equipment and data-networking companies.

A few reports. which are disputed, suggest as many as 38 Huawei employees died from their own hand or from exhaustion during the past decade. Like Foxconn, Huawei experienced horrific on-site suicides, in which an employee typically would throw himself to his death from the balcony of a campus building.

Some commentators have noted that the suicide rate at Foxconn is not inordinately higher than China’s overall suicide rate. Some have even argued that the rate of self-destruction at Foxconn is lower than China’s rate, even going so far as to make the claim that working at Foxconn reduces the risk of suicide for Chinese employees.

Numbers can be sliced and diced, and they can be interpreted in a number of ways. As always, one should verify the accuracy of the source data and carefully check for an inherent statistical bias. I don’t have time to chase that thread now.

So, putting aside that debate, I want to consider another aspect of these stories: the incidence of at-work suicides at both Foxconn and Huawei. The instances of on-site suicide are well documented at both companies.

Perhaps I’m missing something — let me know whether I am — but I don’t believe there ever was a similar outbreak of suicides at technology firms in North America. Cisco, to the best of my knowledge, hasn’t seen its employees leaping to their deaths from the outdoor patios on Tasman Drive in San Jose. I don’t think we’ve seen anything of that sort at Juniper Networks or Brocade — or even Nortel Networks, where people have had considerable reason for despondence in recent years.

Workplace suicide is a dramatic act. It sends a powerful message. The victim makes a statement in not only how he chooses to kill himself but where he chooses to do it.

Ren Zhengfei was right to rack his brain in search of a solution to the morale problem at Huawei. However, as recent events at Foxconn and at other Chinese companies demonstrate, it isn’t a company-specific problem.

As China attempts to move up the technology value chain, from low-cost manufacturing to R&D-led innovation, it will have to find ways of motivating its employees with carrots instead of sticks.

How Brocade Might Connect with Hitachi’s UCP

Hitachi has been said to practice passive — even stealth — marketing. Whatever you call the company’s approach to self-promotion, you’d probably agree that it tends to hide its light under a bushel, at least here in North America, where the company tends to be perceived as an industry afterthought.

That’s why I don’t feel particularly bad about my abject ignorance of Hitachi’s portfolio of networking products, produced through a joint venture with NEC called Alaxala Networks Corporation. Apparently, according to information on Alaxala’s website, Hitachi owns 60 percent of the company and NEC holds the remaining 40 percent.

I was not alone in being in the dark about Hitachi’s status as a purveyor of network infrastructure. Considering that some of Hitachi’s own employees don’t seem to know about this arrangement, I am in relatively good company.

It’s obvious that Hitachi, despite the existence of Alaxala, hasn’t vaulted to the top of the enterprise-networking charts in North America, or in most other parts of the world.

Still, Alaxala could be an important ingredient in Hitachi’s answer to Cisco’s Unified Computing System (UCS) and to HP’s aptly named HP Converged Infrastructure.

Hitachi’s rejoinder to Cisco and HP’s offerings is called the Unified Compute Platform (UCP). It isn’t on the market yet, but it will be released early next year. It will comprise blade servers, storage and network hardware, plus management and orchestration software. Microsoft’s System Center is in the mix, too, as are Microsoft’s Hyper-V virtualization technology and and SQL Server. VMware’s ESX hypervisors also will be supported.

One of the missing pieces is fibre-channel storage networking, but Hitachi representatives, in conversations with technology blogger Nigel Poulton, intimated that the company “might be working” on fibre channel. Then again, as Poulton cautions, that conversation involved significant language barriers, so meaning might have been lost or misconstrued in translation.

As it turns out, the Hitachi Universal Storage Platform V is a key component of the Hitachi Unified Compute Platform (UCP). In that context, it is worth noting that Hitachi already has an existing relationship with Brocade. That relationship involves Brocade providing extensive SAN-switching support for Hitachi’s Universal Storage Platform V.

I think you can see where I’m going here. I’m not the subtlest of characters. There’s a very real possibility that Brocade will be involved with Hitachi’s UCP initiative. To what extent, whether the relationship might be restricted to Brocade’s SAN gear or might also include its Foundry Ethernet switches, remains to be seen.

It’s a relationship worth watching.

At NSN, Nokia and Siemens Still Grope for Exit

Let’s say two companies are involved in a joint venture that’s been an unhappy marriage. The relationship isn’t as toxic as the former partnership between Mel Gibson and Oksana Grigorieva, but it hasn’t been a day at the beach, either. Neither partner wants to remain in the business alliance; they’re both looking for a dignified exit.

With logic and reason as your guides, what would you expect their next moves to be?

Yes, one partner might approach the other, looking to sell its interest in full. It’s also possible that one company might sell its interest to an approved third party, offering a right of first refusal to its JV partner. It’s also conceivable that both partners would put the joint venture on the block, hiring an agent to discreet present it to private-equity shops and strategic buyers. They might even consider putting some lipstick on the pig and trying an IPO, hoping to benefit from auspicious timing and favorable lighting.

Okay, now throw logic and reason to the wind. What would you do now?

Maybe, as Nokia and Siemens have done at Nokia Siemens Networks (NSN), you’d compound the unhappy union by acquiring a floundering telecommunications-equipment business from a vendor eager to unload it. Misery loves company, after all, so why not plunge headlong into the pit of despair? If you put on your absurdist bifocals, the move just might make sense on a surreal existential level. But we’re talking business, not Dadaism.

Just when I think there’s nothing in this crazy industry that can surprise me, something does just that. I admit, I’ve been puzzling over why NSN would buy Motorola’s networks business, which retains some wireless-operator customers, especially in North America, but also carries hefty baggage in the form of a product portfolio predicated on technologies (a large portion of its 3G gear, and its WiMAX 4G offerings) that have gone out of fashion. NSN will pay $1.2 billion for the Motorola unit, and — other than some modest scale and a minor ostensible market-share gain — I don’t see how it derives much benefit from the transaction.

Squeezed from all angles, from traditional competitors Ericsson and Alcatel-Lucent and from hard-charging Huawei — when it’s not fighting an intellectual-property lawsuit launched by, of all vendors, Motorola — NSN isn’t a thriving business. As I have mentioned previously, its joint-venture partners have taken massive goodwill writedowns since forming the business back in 2007.

Digressing for a moment, I want to note that I am not a proponent of joint ventures. Many European companies seem favorably disposed to them, and I understand the underlying reasoning behind them: pool resources, share and mitigate risk, eliminate distraction to one’s core business. Unfortunately, they’re usually unworkable in practice. It’s hard enough getting people from the same company to agree on strategy and to execute successfully. When you have the political machinations inherent in a joint venture, well, the job becomes nearly impossible.

Getting back on track after that brief digressive detour, NSN is in a tough spot.

How tough became clear to me after I read an article in the Wall Street Journal yesterday. Neither Nokia nor Siemens wants to continue participating in the joint venture, but they can’t find a way out. It’s as if Jean-Paul Sartre has rewritten No Exit and staged it in a boardroom. Hell is having to deal with other people in a joint venture.