Yours truly was fortunate to exchange email messages between Aruba and Cisco in an exciting fencing match between the two, regarding the announcement of the new Cisco Mobile Vision.
In the email correspondence published below:
Aruba scoffs at the Cisco announcement calling it markitecture.
Cisco lunges back with how the Aruba argument completely misses the boat and even calls into question Aruba's financial results.
Aruba counter-ripostes that the boat missed was actually the one Cisco failed to catch five years ago when Aruba was formed - namely that customers want mobile applications and that Aruba grew 24% year-over-year in a tough economic climate.
Cisco derobement parry sifts through the rhetoric stating that Aruba financial results are not exactly the mark of a nimble, cutting edge company with the right approach.
Aruba counter-parry, it’s no wonder that Cisco wishes to divert the discussion away from technical prowess in mobility solutions to unrelated matters like Aruba’s stock price (up 9.82% as of market close on June 13).
Cisco parting thrust, we’ll close our participation in this debate by pointing out that market performance is the most important metric of them all. Aruba would love to divert attention from their recent performance, as all of the technical claims in the world mean little without the market vote.
Aruba touché, wired networks are Cisco’s cash-cow and you should always expect them to add ports for seasoning regardless of what WLAN recipe they cook up.
Brad,
Here are some comments about Cisco’s new markitecture.
People move and networks must follow them - that’s the reality of the modern enterprise.
On that point both Aruba and Cisco agree.
How mobility is achieved, and at what cost, is where we diverge.
With this announcement Cisco has at long last accepted that the networking architecture they’ve been peddling for years actually hindered mobility, and they’ve undertaken a total revamp to address the issue.
Their customers should brace themselves for the cost associated with this “upgrade” because it won’t come for free.
Cisco’s new markitecture requires yet another hardware appliance, as well as new licensed software modules, to work with their existing stack of hardware.
All to accomplish what Aruba’s Multi-Service Mobility Controller and vendor-neutral AirWave Wireless Management Suite have done for five years – securely deliver networks and services that follow users wherever they roam.
One justification for the new architecture and new appliance is Cisco’s claim that controllers aren’t capable of processing and disseminating the information required by mobility applications.
That statement is certainly true of Cisco’s own products.
However, Aruba’s Multi-Service Mobility Controllers not only scale to 80Gbps on the data plane, but also include a multi-core processor for scaling the control plane surfaces.
Therefore, for the same services, we provide a more integrated solution than Cisco.
That, incidentally, is specifically why the devices are called Multi-Service controllers.
One need only look at the size and scope of Aruba’s secure mobility deployments to see that Cisco is playing catch-up.
A similar situation occurred in December 2007 when Cisco realized that their vaunted port-centric security and NAC were inadequate for mobile users.
What did they do?
They adopted Aruba’s user-centric, identity-based security model.
In terms of supporting a mobility eco-system, Aruba’s platforms exchange information with other partner applications over industry-standard protocols.
In contrast, Cisco announced the creation of a new proprietary information exchange protocol the objective of which appears to be to lock third party vendors into their new architecture.
Hi Brad – here is our response to the email you were sent…
A nice try, but this response misses the mark, big time...
| 1. | Aruba has no ability to unify networks, they are Wi-Fi only – and mobility is about a consistent experience across wired and wireless (all types). |
| 2. | Aruba has no open architecture for application development, thus requiring any partner to engage with specialized development that results in a drain on engineering and IT resources. |
| 3. | Services delivery has nothing to do with data rate throughput, but rather is all about CPU power and storage ability. |
| 4. | Aruba has no ecosystem of mobility partners to truly enable the applications that enterprises really care about. |
In summary, throwing big numbers (80 Gbps, or in this case, 16 Gbps if one decides to secure their wireless data) as an argument against our approach with Cisco Motion completely misses the boat.
Wireless is deployed not for wireless’ sake, but for the applications.
We believe that building applications that run WITH the network -- not just over them -- requires an evolved architectural approach.
As for catching up, take a look at Aruba’s recent financial results.
Flat to down over the past three quarters.
Losing market share, losing market credibility and losing this particular argument.
In the spirit of the summer movie season... We have spoken, and the critics overwhelmingly agree:
"Its significance is that it's a product that is really separating the network and services layers, and it's open," says Ellen Daley - vice president at Forrester Research.
"This means [software] developers, and groups within Cisco, can leverage or use network resources more easily."
"The combination of service programs plus the API is an attempt to link existing context and security infrastructures with an emerging idea of enterprise-wide mobility," says Maribel Lopez - CEO of Lopez Research LLC, a technology research firm.
"I think of this [announcement] as bridging together disparate islands first," she says.
"First you need to connect and enhance existing applications, then once you have that bridge, you can offer new services."
Michael Morisy - SearchNetworking.com:
"It's a whole new way of looking at provisioning applications on networks," said Craig Mathias - principal analyst with the Farpoint group.
He said that Cisco is trying to take components once handled at the application level and bake them deeper into the network itself in order to provide better contextual information and make it more widely accessible to third-party developers.
"I think their statement of direction about where they are headed with mobile computing is crucial for their customers," says Ken Dulaney - vice president of mobile computing for technology research company Gartner.
"While there's a lot more to do, it's a very good first move by Cisco."
Brad,
To paraphrase a famous Roman statesman, Cisco in general is quick to believe that which it wishes to be true.
We welcome Cisco’s interest in mobility solutions but question why they believe they’ve invented a new category when in fact they’re just late to an ongoing party.
It’s also surprising how little Cisco appears to know about a solution that each day makes converts of dissatisfied Cisco customers.
Let’s address each point in turn:
| 1. | Aruba is hardly Wi-Fi only – our solutions encompass wired, Wi-Fi, enterprise fixed mobile convergence, and remote access solutions.
Our vendor-neutral AirWave Wireless management solution manages Wi-Fi, WiMAX, metro mesh, point-to-point networks from more than 16 vendors – Cisco included. Aruba offers unified mobility solutions that include multiple media, technologies, and vendors - that’s a feat Cisco simply can’t match. Cisco is also not in a position to throw stones about wired and wireless integration: in Gartner’s 2007 Magic Quadrant report analysts Michael J. King and Ken Dulaney caution that “Cisco continues to have split wired and wireless management products; prospects for integration in the coming years are slim.” |
| 2. | If Cisco’s point was to highlight that proprietary protocols are anathema to open application development, then we agree with them.
Strange, then, that Cisco is developing a proprietary MSE-to-Cisco-only NMSP exchange protocol as their solution. NMSP itself is replacing LOCP, an Airespace proprietary protocol. Aruba uses SNMP for management and an open API layer based on XML and other open standards. |
| 3. | Reliable service delivery encompasses a range of factors of which data throughput is one important element.
Does Cisco espouse streaming continuous video over a low speed Bell 202 modem? Of course not – a network must be designed for speed. We agree that CPU power is also important, which is why we use multi-core processors and hardware-based encryption accelerators. Storage should be appropriate for the intended application, not an end unto itself. And based on lessons learned from the storage industry, long-term storage should be kept in devices dedicated for this purpose and not in controllers. |
| 4. | Aruba has an active and fast growing technology partner program with more than 26 members, including companies such as AeroScout that we share in common with Cisco.
Since Cisco is unaware of the program we’ll give them a little help by printing the partner program URL: http://www.arubanetworks.com/partners/solution_partners.php |
The boat that was missed is the one Cisco failed to catch five years ago when Aruba was formed - namely that customers want mobile applications.
The quotes from journalists and analysts aren’t vindications of Cisco’s strategy - rather they express relief that Cisco has at long last finally declared the failure of their current architecture with respect to mobility, and created an entirely new design to correct this oversight.
Rest assured that this news won’t be as welcome to Cisco’s customers, who will pay dearly for yet another layer of hardware and licensed software to accomplish what Aruba has built into its controllers for years.
Aruba’s standard controllers enable role-aware wired / wireless networks AND they integrate with any wired IDS and NAC solution in the market without added cost.
If the parting shot about financial stability was intended to be a body blow, Cisco erred again – they looked at the wrong numbers.
Aruba grew 24% year over year in a tough economic climate.
A quick perusal of our announced design wins shows a vibrant, growing enterprise.
Aruba from bondage will deliver Cisco’s customers, to paraphrase Shakespeare, and we’ll do so from the position of a best-in-class supplier with a healthy balance sheet and outstanding support services.
Hey Brad - happy Saturday to you... Sifting through all of the rhetoric here, we boil this down the following:
| 1. | Customers invest in wireless networks to enable mobility applications.
We believe that the network architecture simply must evolve to deliver integration, simplicity and ongoing greater value for customers. Opening up the network platform for open application integration is fundamental to our approach, not only with our wireless portfolio, but across other Cisco offerings. We clearly disagree with Aruba, and do not believe the WLAN controller is the element to open up for the ISV community to develop mobility applications that work with the network. Such an approach does not scale, and inappropriately provides open interfaces to what is a tightly secure network element. Cisco has recognized that an evolved approach is needed in our industry, not just because we figured it out on our own, but because our customers (Cisco customers, Aruba customers alike) constantly push us in this direction. And they do so because they believe we are the only vendor capable of achieving this level of integration to make the entire mobility ecosystem work better. It is worth noting that in the hundreds of discussions with industry analysts and experts leading up to the launch of Cisco Motion, the name Aruba Networks was rarely raised by these experts. Cisco Motion addresses key issues facing mobile devices, the network, the applications, and most importantly, the end users who enjoy the experience, and the IT departments who need a better experience in making it all happen. The debate being offered up here by Aruba is inherently limited to the WLAN "box" level. Again, the "get-it" factor is missing. |
| 2. | Market results (and recent ones at that) are the best scorecard on who has the right approach.
Aruba's financial performance has slowed to a crawl over the past three quarters. The company's market cap has collapsed, disproportionately so, to the overall market. It's almost certain that you will see this reflected in Aruba's y/y growth rates which will slow significantly -- reflecting the wall the company appears to have hit with its previous-generation, limited approach to enterprise wireless networking. One look at Wall Street projections tells the story. Wall Street analysts expect Aruba's y/y growth to slow to 11 percent this quarter, and to 5 percent in the following. You can hear the tires screeching to a halt. Not exactly the mark of a nimble, cutting edge company with the right approach. |
Brad,
We agree that readers should listen to what the analysts have to say about Cisco Mobile – no need to take Aruba’s word for it.
Analysts have noted that Cisco has been conspicuously absent from the enterprise-mobility marketplace, and that Cisco Motion points to where Cisco believes it can provide value and define a full mobility story where none existed previously.
As stated in our last salvo, Cisco is about 5 years late to the mobility party.
With regard to the function of the controller, Aruba packs more functionality into a more powerful and smaller platform than Cisco offers.
To provide the same wired and wireless network functionality as a single fully populated Aruba MMC 6000 Multi-Service Mobility Controller (drawing a total of 680 Watts), Cisco requires 5 fully loaded Catalyst 6500 chassis (drawing more than 10,000 Watts not including Cisco’s newest Mobile boat anchor).
The standards-based mobility interface described in our last response operates through the multi-vendor AirWave Management Suite, not the controller, so Cisco’s concerns about controller security are moot.
That said, Cisco does raise a red flag about controller level application layer intelligence – namely that it’s essential to have this capability within the network in order to provide seamless mobility.
Wireless LAN data path services can be managed by low-scale controllers, and wireless LANs can be deployed on a large-scale simply by installing more PC servers in the data center, but only when the wireless LANs are a secondary form of connectivity.
That is clearly Cisco’s position since they don’t have networking intelligence built into their wireless LANs – and they don’t appear ready to add it any time soon.
Aruba believes in, and is today delivering, the all-wireless workplace in which 802.11n wireless LANs are the primary form of connectivity.
And that objective has informed our controller architecture:
 |
Stateful firewall security services preventing the need to upgrade existing firewall infrastructure when moving from wired to wireless; |
|
Role-based policy enforcement enabling the scalability of mobility infrastructure without compromising security of the internal network; |
|
Identity-aware policy enforcement that is easy to integrate with any NAC solution, enabling secure mobility without hindering existing wired security infrastructure or requiring additional ‘in-line’ NAC appliances for every deployed WLAN controller; |
|
Fast and best-in-class security for 802.1x services through the use of centralized encryption; and |
|
PBX-agnostic enterprise FMC solutions integrated within the data path (where else would you want it?). |
At some point in the future, when Cisco actually embraces the mobile enterprise, these features will start appearing on their controllers.
Until then, it’s wired networks first (Cisco’s cash cow), wireless LANs as poor second cousins in Cisco’s new markitecture.
Interestingly, Cisco is conspicuously silent on the subject of network management, an Achilles heel for them today and one that presumably will get worse as partner devices are added in the future.
Aruba offers the only vendor-neutral management platform that supports products from more than 16 vendors.
Cisco’s management software is proprietary, limited to their own closed systems, and severely lacking in features, functionality, and ease-of-use.
All service level storage, monitoring, troubleshooting functions offered by the Cisco MSE are already, or planned to be, in place on the AirWave platform:
|
Multi-vendor WLAN infrastructure management; |
|
Device agnostic client level troubleshooting and monitoring; |
|
Open XML API for WLAN information; |
|
Integration with RFID systems; |
|
Client location tracking with historical physical mobility patterns of client devices without requiring additional licensing, server or manual RF fingerprinting; |
|
Voice over WLAN reporting and monitoring; |
|
3rd party WLAN configuration management and automatic configuration auditing; and |
|
Advanced wireless IPS functionality. |
Given the foregoing, it’s no wonder that Cisco wishes to divert the discussion away from technical prowess in mobility solutions to unrelated matters like Aruba’s stock price (up 9.82% as of market close on June 13).
This humorous maneuver is classic Cisco, and snookers neither readers nor prospects.
One presumes that Network World readers are more interested in topical IT issues than the financial opinions of Cisco’s PR department.
I know we are.
G’morning Brad,
Aruba is still not getting it — continuing to argue Wireless LAN box features because this is the limit of the company... an overlay WLAN vendor.
For reasons we have already stated, the Cisco Motion strategy is far more encompassing than WLAN controllers, because mobility is about far more than a WLAN.
We can continue to do data sheet feature back and forth, where we are very confident in our ability to competitively line up Cisco’s WLAN capabilities with Aruba’s, but we’ll take that particular argument directly to the market.
We’ll also let Aruba try to figure out where our broader play with Cisco Motion is going on its own (hint — it’s not about the box/appliance/controller).
For the readers, feel free to visit www.cisco.com/go/wireless for more information.
We’ll close our participation in this debate by pointing out that market performance is the most important metric of them all.
Aruba would love to divert attention from their recent performance, as all of the technical claims in the world mean little without the market vote.
With all of the bravado displayed in the last post, one must ask how Aruba’s vaunted WLAN offerings are actually performing in the market of late:
| A | According to the Dell’Oro Group, a noted market share firm recognized both by the networking industry and Wall Street alike, Aruba’s total Enterprise WLAN marketshare has slipped from 9 percent to 7 percent over the past four calendar quarters.
Also according to Dell’Oro, Aruba revenue garnered by WLAN controllers and access points has gone from $32 million four quarters ago to $26 million in the most recently reported quarter. Dell’Oro shows that Cisco has gained over this time period. |
| B | In the same Dell’Oro report, the WLAN controllers that Aruba spends a lot of time boasting about in the previous post have gone from $22 million revenue and 16 percent market share four quarters ago, to $14 million and 10 percent market share in the most recently reported quarter.
Dell’Oro shows that Cisco has gained 6 points of share in this time period. |
| C | In the emerging 802.11n next-generation market, where Aruba has boasted that they will make significant inroads against Cisco, the current Dell’Oro group revenue market share has Cisco with 71 percent share in 802.11n access points (greater than our overall WLAN market share).
Aruba, trails with 4 percent. |
Aruba is doing a lot of barking here, but the market reality takes away most of their bite.
They’ll most likely respond with how they have grown over the past 3-4 years.
This fine — startups that go public do indeed grow.
But, the past year is a telling sign.....
Brad,
We are pleased to hear that Cisco has posted its last entry because their drum beat is both off-time and monotonous.
The technical merits of Aruba’s response remain unchallenged by Cisco and the reason why is no secret – Cisco’s mobility solution comes up short.
If Cisco wants to stand on the merits of the Cisco Mobile markitecture then they need to pony up the "data sheet" level details they deride Aruba for providing.
All we have from Cisco is a gentle breeze from hand waving and little else.
The harsh reality is that it will take 12-18 months before the details of Cisco’s markitecture make it to market, longer before there is any semblance of system-level integration.
Aruba offers unified mobility solutions – complete with data sheets – today.
Cisco states that mobility is about more than wireless LANs, and we agree.
That’s why Aruba’s solutions encompass 802.11n wireless LANs, wired LANs, cellular integration, and remote access.
That said, customers – not vendors – will decide what technologies best address their mobility needs.
In a recent press release Microsoft stated that more than 75 percent of their employees use the Aruba wireless LAN every day, 72 percent said they could work without any wires at all, 93 percent can use their computers in new locations because of the enhanced mobility afforded by the Aruba wireless LAN, and 70 percent of employees said the Aruba network saved them at least 5 work hours a week because of increased flexibility.
Wireless LANs are not Aruba’s only mobility solution, but customers say they have a central role to play and we agree.
The significant performance and integration limitations of Cisco’s own WLAN technology probably taint their view of this valuable resource and limit their ability to exploit its benefits.
Also, wired networks are Cisco’s cash-cow and you should always expect them to add ports for seasoning regardless of what WLAN recipe they cook up.
Aruba has no such fetters.
Turning to Cisco’s tunnel-vision focus on Aruba’s first quarter performance, we have only one comment – yawn.
A lot has transpired since then and it’s not good news for Cisco.
In the intervening time Aruba shipped its 10,000 802.11n access point, which we estimate amounts to more than half of all enterprise 802.11n access points shipped by the entire industry in that period.
Given that we did not start shipping in volume until the end of the first quarter, that’s an important milestone.
Aruba thanks you for the opportunity to participate in this interactive discussion.
Your forum informs readers about technical issues and marketing messages in a way no article can replicate.
What do YOU think of Aruba scoffing at the new Cisco Mobile Vision announcement as being markitecture?
About Brad Reese on Cisco
Brad Reese cofounded BradReese.Com Cisco Refurbished which offers one year warranties on Cisco Refurbished and Cisco Repair.
Contact him.
Brad's blogroll
Brad Reese on Cisco archive.
Cisco Subnet
The opinions expressed in this Weblog are those of the writer and may not represent the opinions of Network World.
|
|
Aruba vs Cisco
This dust-up between Aruba and Cisco is amusing.One part of the Cisco announcement particularly caught our eye. The announcement from Cisco has validated the message that AirTight has been offering for some time and which Aruba validated a year ago with is purchase of NetChem. That is, there is a need for infrastructure providers to enhance mobility solutions. This move specifically reinforces the importance of wireless intrusion prevention (WIPS)as a necessary enhancement to the infrastructure. The debate for companies concerned about wireless security now shifts from, ‘Do I need a wireless intrusion prevention system (WIPS)?’ to ‘What system will provide the most functionality and security while reducing my total cost of ownership?’ Cisco’s announcement reinforces our contention that WIPS is required to secure your network from wireless threats and companies should be looking for the best in class and evaluating all solutions.
This announcement leaves no doubt that Cisco has heard its customers and understands that every customer planning to run any business critical applications on a wireless network needs a robust, scalable wireless intrusion prevention system. Cisco’s new architecture calls for deploying sensors as an integral part of building this wireless mobility platform. While this validates the need for sensors it does not address other challenges. WIPS providers like Airtight have evolved UI and technology solutions which are specifically aimed at reducing operational and management costs.
The market, however, still does not know if wireless security is something an enterprise can manage in house or something it will eventually need to outsource to those with specific expertise. Cisco’s architecture will force users into limited options while the jury is still out on this question.
We are also happy to see Cisco providing an open platform for integration with third party vendors. AirTight has led the WIPS industry to offer API level integration with Cisco’s WLSE and Aironet IOS APs as well as API level integration with the Cisco wireless LAN controller (WLC) and lightweight APs in our SpectraGuard Enterprise 5.5 release. This move opens up many more possibilities for future integration making interoperability with Cisco infrastructure even easier in the future.
Aruba misrepesentations
Reading through Aruba's diatribe here reinforces my notion that you really can't trust what this particular vendor says.
This company announced their 802.11n products in early November, and here admits that they never started really shipping until April. Hmmm ... nimble technology leader announces a next-generation product and ships 5+ months later? Bizarre marketing strategy, or lack of execution?
But it doesn't stop there -- there is a key detail behind those claimed 10,000 units. Aruba offers APs that can be software upgraded in the field to 802.11n in the future. So again we have a deceptive claim. Vast majority of those APs aren't operating as 802.11n APs.
Lastly -- for all of the bragging about their next-generation controllers (announced in early November), there is scant evidence of product shipping of this platform. Another case of lack of execution? The analyst numbers Cisco cites above showing Aruba's shrinking controller business seem to illustrate this clearly.
Technology leadership comes in many ways -- but most importantly, actually delivering to market should be the key filter here between press release hype and true leadership.
Aruba's huge quarter miss makes a lot more sense now.... pre-announcing next-generation products and not shipping for 5+ months. Either a lousy marketing strategy or clear case of engineering lack of execution. In any case -- it's hard to know what to believe when Aruba speaks...
Official Aruba Response
Readers wishing to read customer’s own words about Aruba’s market-leading 80Gbps MMC-6000 Multi-Service Mobility Controller, AP-124/AP--125 802.11n Access Points, and AP-124ABG/AP-125ABG Field Upgradable 802.11n Access Points need only point their browsers to:
http://www.arubanetworks.com/company/press_release
Those considering an 802.11n deployment may wish to know that Aruba’s 802.11n access points weigh only 11 ounces (Cisco’s cinder block weighs nearly 5 pounds), are only 1/3 the physical size of Cisco’s 802.11n AP (see drawing below – Cisco’s AP is represented by the dashed red line), and offer full dual-radio 3x3 MIMO operation powered from a single 802.3af power supply (Cisco requires a hefty high power 802.3at source).
Works great, less filling, more thrifty – that’s what customers want.
On the subject of power consumption, Aruba’s MMC-6000 controller draws only 680 Watts when fully loaded, while Cisco requires 5 fully loaded Catalyst 6500s to provide comparable features.
Assuming that electricity costs $0.10/kWh an Aruba user will save (and a Cisco user will be penalized) $8,800 per year in electricity costs, $3,000 per year in HVAC loading costs, and $6,350 in maintenance costs.
Lower OPEX – that’s what customers want.
Aruba is positioned as a Leader in Gartner’s Magic Quadrant for Wireless LAN Infrastructure 2007 by Michael J. King and Ken Dulaney, dated 20 December 2007, which notes that Aruba has a strong technical vision that is well-implemented, and offers best-of-breed management and control functions.
A visionary leader – that the type of company with which customers want to do business.
Thank you again for the opportunity to participate in this forum.
Sincerely,
Brad Reese
http://www.BradReese.Com
3x3 MIMO. Does that mean it
3x3 MIMO. Does that mean it can handle 3 spatial stream MCSs? Atheros chipset only supports 2 SS.
Is it dual mode?
Here's some irony.
The 50+ Cisco employees at Interop were using a wireless network donated by Aruba.
Size/Power Matters
LOL!
My AP is smaller than your AP? My power is stronger than your power ? Sounds like two kids on the playground or two men in the shower comparing toys? Boys will be boys?
In this case, Aruba is right ,the smaller the better, the less power the more efficient with energy
Cisco is right too. Aruba is an island... Cisco is a continent.
Why can't we be friends?
MERGER?
Post new comment