When I first joined SearchNetworking.com I ran a survey asking you what the No. 1 security threat was in your industry. Your responses impressed me; even to this day I get haunted by the survey results: Across every vertical, a wild majority of networking professionals considered their own users to be their No. 1 network security threat.
Although these were findings from three years ago, the one thing that hasn’t changed on the network is that people are still on it. No matter how technology changes, your users will be there. They necessitate a network, yet break the security of it. Why? There are all sorts of reasons, some of which you can resolve.
How you ask? Our network user management guide explains. In it we highlight several problem user scenarios, from the make-it-so CEO to the Internet novice who doesn’t know the difference between the Internet and their website (yes, that was a The Website is Down reference). In these user profiles we determine their network threat level and how to mitigate their bad network behavior. But nothing could cover every user scenario. If you have a way of dealing with the 8th layer of the OSI model — people — let us know!
This is a story of a small-time vendor who, after years of being stepped on by a billion-dollar analyst firm, throws open the windows of its San Jose office and screams, “I’m mad as hell and I’m not going to take it anymore.”
The question is, will anyone else join them?
ZL Technologies, an email archiving company who has languished in the niche corner of Gartner’s Magic Quadrant for four or five years, is suing the analyst firm for $1.6 billion, claiming defamation, trade libel, unfair competition and negligent market interference, etc.
Gartner contends that it has a First Amendment right to express its opinions on technology markets. It’s hard to argue with that.
As I have mentioned before, there are many people who dislike Gartner and its influence over the IT industry and claim that its Magic Quadrants are more about who you know and what you spend with Gartner than the quality of your product and your company.
My colleague Mark Fontecchio did a nice job of collecting some reaction to this lawsuit from the people who matter: IT decision makers. It’s clear that for all the ridicule that bloggers, out-of-favor vendors and competing analysts heap on the Magic Quadrant, the market measuring tool still has significant influence over buying decisions. One IT veteran said the Quadrant represents about 20% of the decision-making process while another said a CIO should always review a Quadrant with his CEO before buying a product.
IT executives have a fair amount of skepticism about the Quadrant but they still rely on it to some degree. Every time I write about one, the page views come pouring in. Cisco might be the leader for the 100th time in a row for enterprise Ethernet switching or wireless LAN, but everyone still wants to know the nitty gritty about the Quadrant.
So what could Gartner do to reduce criticism? Well, how about more disclosure? Exactly how does that magic box get drawn up? Maybe the mathemtaical model should be disclosed. How much money does each vendor on the Quadrant spend with Gartner? How many hours has the analyst spent with each vendor? How many customers of each vendor were consulted? These are facts that inquiring minds want to know.
Juniper Networks is planning a huge technology announcement next week. So big that the company has invited press to a gala Wednesday night and then a day packed with a press conference in the morning and executive break-out interviews all afternoon Thursday.
The company has pre-briefed every analyst in sight and clearly plied them with drinks, money or … something … because none of them will drop the dime on the announcement either.
What is clear is this: This announcement had better be data center related. After all, last February, Juniper announced it was working on the Stratus Project – a data center initiative meant to combine storage, compute, switching and networking all on Converged Enhanced Ethernet (CEE). But the announcement included no specific product details or actual release dates.
It appeared that Stratus was a slap back at Cisco, which “leaked” news of its upcoming blade server and Cisco Unified Computing data center architecture just a month earlier.
The thing is, Cisco followed swiftly (in March) with actual product announcements and release dates.
Juniper’s non-announcement frustrated those in the networking community so much that one reader on NetworkWorld’s Cisco Subnet Blog, wrote in:
“Juniper doesn’t have anything or they would be releasing it! They got caught holding their imaginations and have no where to turn. ‘I have a super special architecture that will crush all others, will blow your mind, run all OS’s, run terabit Ethernet, and make coffee in the morning. Due out 2013.’ Give me a break, its like little children on the playground.”
That may have been a bit harsh, but it’s time for Juniper to come forth with a solid data center plan if it really wants to be competitive. Since Juniper said that Stratus was meant to be an architecture that would be built on a number of vendor partnerships in combination with its existing technology, this very cryptic Juniper press invitation does lead one to believe the hoopla is, in fact, about the data center after all.
“This will be the company’s most significant event in their 12-year history and will be hosted by Juniper’s CEO, Kevin Johnson. The company will be unveiling new hardware and software products, new partnerships and new go-to-market programs as part of the event.”
Juniper set the stage, now it’s time to put on the show. We’ll be there to let you know if that actually happens this time around.
As a follow-up to my post on analyst integrity, here is a rather visceral rant against Gartner from David Amsler, founder and CIO at ForeGround Security, an infosecurity consulting and services firm. Amsler’s post is a somewhat cheeky response to Larry Chaffin’s evisceration of Gartner at Cisco Subnet posted back in August.
Amsler’s main argument is that Gartner is an arrogant organization that brings little value to the IT industry. Well, his words are actually, “they truly don’t provide any real value to our industry.” For proof of arrogance, he points to this privacy debate which broke out on Burton Group analyst Bob Blakeley’s blog. He also deconstructs Gartner’s ballyhooed Magic Quadrants, calling Gartner’s method for constructing them “fishy.”
I’ve never had a problem with Gartner, myself. I have had run-ins with vendors who were unhappy with Gartner (and my coverage of Gartner’s research). I find that most Gartner analysts — at least those who return my calls — are as knowledgeable as any other subject matter experts I talk to on a daily basis.
Since I’m a member of the press, people aren’t always candid with me about their views on subjects like this. Or their candor is issued to me under the blanket cover of off-the-record comments, which means I can’t share them with anyone. In any case, I remain curious about this whole issue and I hope to hear from readers who have an opinion (whether on or off the record).
Gartner VP and distinguished analyst Tom Bittman wrote a passionate rant on his blog yesterday defending his and his coworkers’ integrity as analysts.
In my 14+ years at Gartner, I have never, ever allowed a vendor to influence my opinion with anything but facts. Period. They have certainly tried to influence me with non-facts. I can say this definitively – it has never worked.
Tom’s rant has sparked an interesting debate among current and former analysts, bloggers and vendors in his blog’s comments section. It’s worth reading.
There has been plenty chatter in the blogosphere and elsewhere about whether analysts are offering any value to enterprise clients. Many analyst firms, such as Gartner, count both IT vendors and end user enterprises as customers. It’s hard to serve both masters while maintaining an appearance of integrity. This is probably why Gartner has an ombudsman... the only ombudsman I know of in the analyst industry.
Gartner’s use of an ombudsman doesn’t protect it from attacks on its integrity. James Watters, for instance, suggested recently that IBM’s favorable position Gartner’s Magic Quadrant for Web Hosting and Hosted Cloud System Infrastructure Services was due to IBM being a very important paying client of Gartner. Trapeze Networks came to me after I wrote about Gartner’s last wireless LAN Magic Quadrant and suggested that Gartner had unfairly placed Meru Networks ahead of Trapeze because Meru’s current VP of marketing is a former Gartner analyst and one of the authors of the MQ was very briefly an employee of Meru.
It’s hard to prove your integrity. For some people, there just isn’t enough proof in the world to erase any doubts. Journalists face the same challenge. In my newspaper days people would occasionally accuse me of being a mouthpiece for politicians. I knew I was doing my best to write fair and balanced stories, but the accusations still stung.
Gartner’s use of an ombudsman is significant, I think, although the office would have more impact if the ombudsman’s blog were updated more than once a month. An ombudsman is the voice of the client (or in a the traditional journalism sense, the reader). Clients will have more faith in a an analyst’s advice if the firm has someone on staff who tackles each and every significant attack on its integrity head-on, in full view of the public.
Gartner and other analyst firms should also disclose the nature of their paid relationships with the vendors that they evaluate for enterprise clients. This way, clients can understand the full context of the advice they are receiving. Some firms do this. Others don’t.
If you could talk to the CEOs of the leading analyst firms out there, what advice would you give them to help them establish that they are giving unbiased and independent advice to enterprises?
Oracle CEO Larry Ellison dismissed the Wall Street Journal’s speculation that his company (among others) was interested in buying Brocade while speaking at the comapny’s annual shareholders meeting yesterday, according to MarketWatch’s John Letzing, IBM, HP and Juniper remain potential suitors according to the Internet rumor mill.
Meanwhile, Extreme Networks’ stock price took a relative beating on Tuesday after it lowered its earnings for its first fiscal quarter (which ended Sept. 27). Wall Street analysts were expecting Extreme to make $80.4 million but Extreme says it will be closer to $66 million. Extreme cited supply chain problems, which should scare customers. Other networking vendors have admitted to supply chain problems in the wireless LAN market, but I haven’t heard anything about wired networking supply chain problems before. Extreme said it has lost some deals and had other delayed by the problem. Are Extreme’s suppliers giving preference to larger networking vendors?
Extreme’s market capitalization is at $231 million as I write this. It would come a lot cheaper than Brocade and would offer someone like HP ProCurve a nice complementary line of data center switches.
A year ago Brocade was a buyer. Now its a seller and HP or Juniper might be buying.
Citing unnamed sources, The Wall Street Journal reported yesterday that Brocade Communications has put itself up for sale. This news comes less than a year after Brocade closed on its own acquisition of Foundry Networks, a $3-billion deal which gave the Fiber Channel storage networking market leader a broad portfolio of high-performance Ethernet switches and a variety of other data center networking products. Brocade has since inked some high profile OEM agreements with companies like IBM, creating a promising sales channel for the Ethernet products.
The Journal reported that HP and Oracle were likely suitors for Brocade. Oracle seems like a strange fit, but HP makes sense, given that Brocade’s former Foundry switches would fill a gap in HP’s ProCurve switching portfolio. ProCurve lacks a track record in high density Ethernet switches, something Foundry specialized in. The deal would also boost its storage business and give HP a much broader overall suite of data center products.
Now Journal blogger Michael Corkery speculates that Juniper might be a buyer for Brocade, too. Known more for its service provider routing business, Juniper recently rolled out its own line of enterprise Ethernet switches, but it has $2 billion in cash. A Brocade acquisition would give Juniper a much more established switching business, plus a lucrative storage networking business with deep and successful partnerships with IBM and EMC. But Juniper’s stated switching strategy is centered around the concept of having one operating system, JUNOS, across all of its networking products. Buying Brocade would force a complete change in direction.
Cisco can go one of two ways with its $3 billion Tandberg acquisition: It can get over old habits and work toward interoperability in videoconferencing and telepresence — or not.
The overarching goal of the Tandberg acquisition was for Cisco to buy its way into the mass video conferencing market with Tandberg’s lower-end, high-definition technology to add to its own high-end telepresence offering. But Cisco CEO John Chambers also noted that Tandberg is “good at” working with other companies, and that Cisco would ultimately offer multi-vendor interoperability.
Cisco doesn’t exactly have a great track record when it comes to interoperability. If video systems are built on the H.323 protocol or Session Initiation Protocol (SIP) they can interoperate. Both Tandberg and rival Polycom use H.323 and have worked together on standardization bodies to establish industry-wide interoperability. Cisco, on the other hand, has been blatant about its disregard for interoperability in videoconferencing (along with lots of other technologies).
Last May when I interviewed Cisco about telepresence interoperability, a senior official explained the company’s stance:
“Standardization and innovation sometimes work against each other,” David Hsieh, senior director of marketing for Cisco, said. “Interoperability is the lowest common denominator.”
Hsieh added, however, that Cisco could change that view with pressure from customers. Maybe the pressure has arrived. Or maybe Cisco just figures that bringing video to the masses won’t happen by locking customers into one brand that can only be used within their own company or with other participating customers.
Still there is no promise that Cisco will suddenly begin working toward standardization and interoperability. First, most of its networking equipment remains proprietary. Cisco’s other collaboration applications are not easily interoperable with those from other vendors – often creating islands of collaboration rather than a mass system. Plus, it will take time for Cisco to integrate Tandberg’s technology with its own. When that happens, standardization could get lost in the shuffle.
Ultimately, the future of the videoconferencing market – which has had numerous false starts and sudden lows – depends on what Cisco chooses to do here.
Two weeks ago HP ProCurve launched a new line of blade switches, the 6120XG and the 6120G/XG, which plug into the same blade enclosures as HP’s BladeSystem servers. These switches were meant to replace the switches which HP OEMs from a third party, most likely Blade Network Technologies’ BNT switches.
Now Cisco has launched the Nexus 4000, a blade switch which also fits inside a blade server enclosure. Like other switches in the Nexus family, the Nexus 4000 runs the NX-OS operating system, which Cisco designed for converging LANs and storage-area networks (SANs). The switch is also designed to support converged, data center Ethernet, the pre-standard technology that will allow data centers to run server and storage traffic across the same wire.
Although the Nexus 4000 is part of Cisco’s Nexus switch family, Cisco is only selling it through OEM channels, specifically blade server vendors. These are the same server vendors that Cisco now competes against with its Unified Computing System, a line of data center technologies which includes blade servers. It remains to be seen how enthusiastic vendors like IBM and Dell will be about selling a switch from a server competitor with their blade server systems. I doubt HP will be interested, given that it just launched a similar product.
It is ironic that right as Gartner released research showing engineers are managing 20% to 30% more network components than they were last year, networking pros are about to receive even more on their plates by way of facilities management … of all things.
This week the IP for Smart Objects (IPSO) Alliance announced a certification and compliance program to ensure vendors release IP-based smart objects that interoperate. You may be thinking, what the h&%ll do I care about smart objects? What is a smart object?
Believe me, you should care. Smart objects are sensors and actuators that run in a range of applications from smart grid management to building automation. The IPSO alliance is hoping to accelerate the use of IP in these devices. That means that enterprise lighting systems, for example, can be networked into the LAN. Among the many use scenarios, facilities management systems will communicate to network managers when a system is wasting power without actual use. It is feasible that these systems could save enterprises as much money as greening data centers.
On a larger scale, IP-based energy and building management means that utilities will implement smart meters in enterprises and homes that will warn home base of outages and waste. These smart meters will be part of larger IP networks laid over utility grids. In addition to enterprise and consumer savings, regional governments will spend less, enabling them to direct money to other crucial projects.
Cisco has already moved in on the action with its EnergyWise software initiative. EnergyWise is embedded into Catalyst switches and used to control not only power within the network, but also on PCs and building control systems. So it’s no surprise that Cisco is a member of the IPSO Alliance, along with Ericsson, SAP, Sun Microsystems and Google.
All of this innovation may be exciting, but for network engineers it means the need to gain more skills (possibly even certifications), more responsibility and assuming an even greater role in every enterprise. As this happens, they will have to demand the manpower and support necessary to expand in this new direction.