x
Optical/IP

Cisco Prowling WAN Optimization?

Now that Juniper Networks Inc. (Nasdaq: JNPR) has coughed up nearly half a billion dollars for WAN optimization player Peribit Networks Inc. and traffic management vendor Redline Networks Inc., the startup world is abuzz with the potential for a response from Cisco Systems Inc. (Nasdaq: CSCO).

WAN optimization and traffic management technology were together seen as something of a niche market, prior to Juniper's deals (see Juniper Takes Two: Peribit & Redline). Now they appear to be the focus of every investment banker between Menlo Park and Manhattan (see Peribit Deal: More to Come).

In fact, according to most sources in the space, one would think it's a foregone conclusion: Many informed onlookers consider it inevitable that Cisco is going to pull the trigger on at least a WAN optimization company -- and possibly a traffic management company as well.

The reasoning? Many technology experts see the application delivery and management space as heating up, and it could pose a challenge for Cisco’s current product portfolio (see Zeus: A Takeover Target? ). This conclusion was being expressed by many experts in the space at last week's Interop tradeshow.

For example, Eric Klinker, vice president of engineering at service provider Internap, believes that Cisco doesn't have the right WAN optimization tool, and will likely launch a counterattack to Juniper's $337 million acquisition of Peribit. He sees the most likely candidates as Expand Networks Inc., Packeteer Inc. (Nasdaq: PKTR), and Riverbed Technology Inc., but he considers Riverbed the "thought leader."

Klinker, who describes his company as a "Cisco shop," says Cisco hasn't had much success in building it's own WAN optimization technology, which involves heavy computing of algorithms to more efficiently compress bandwidth used by applications. That's what led Internap, a service provider, to buy NetVMG, where Klinker served as CTO.

According to Klinker, Cisco tried to develop WAN optimization functions on a router as part of its Optimized Edge Routing (OER) technology, but it didn't work, because the heavy computing power of WAN optimization needed a separate device.

"You cannot put that function on a router," he says. "It fried the router -- it was not a successful launch at all."

When contacted on Friday, Cisco was unable to locate anyone to address Klinker’s specific concerns about OER. Version 1 of OER, which works in conjunction with Cisco’s IOS software and WAN edge routers, became available last year. The company is still touting OER as a way to help network administrators minimize bandwidth, and the technology aims to distribute packet loads across the network and detect failed data paths at the edge of the WAN. The vendor also offers an OER appliance to host the software application for traffic load distribution and failure detection.

Baruch Deutsch, director of product marketing for Cisco's caching division, sought to downplay any suggestions that the vendor is a WAN laggard. “We have a wealth of products in the WAN optimization space, and we will continue to develop that,” he says.

Deutsch pointed to Cisco’s acquisition of Actona last year as evidence of the firm’s commitment to this space, although he was less forthcoming when asked about any future M&A plans (see Cisco Completes Actona Buy). “I am not going to comment on Cisco acquisition plans,” he says.

Meanwhile, the WAN optimization startups are basking in some newfound glory. Ariel Shulman, vice president of business development at startup Expand Networks Inc., says his company has definitely taken notice of the Juniper deal for Peribit.

"It pushed everything one league up, put this WAN optimization thing on the map," says Shulman. “Right now we have stopped evangelizing and started competing in a market that is validated.”

When asked if Cisco needs a WAN optimization tool and would look at buying a company such as Expand, Shulman says, "There is a gap [in their portfolio], I could agree."

Expand now has more than 1,000 customers, according to Shulman. It has raised a total of $40 million, and its last round of $10 million is being used to expand its salesforce.

One reason the WAN optimization startups are so excited about Juniper's deal is the valuation. Juniper paid a combined valuation of 12x sales for both Peribit and Redline, according to some recent research reports. To put the deal in perspective for Peribit, the startup had raised about $40 million in venture capital, so the purchase price of $337 million represents a hefty gain for investors.

In addition to putting focus on the WAN optimization space and Peribit, the Juniper deals have also breathed new life into traffic management, where the acquisition game is largely believed to be in play.

"We have long expected Cisco to go into this market," writes Steve Kamman, a CIBC World Markets analyst, in a recent research note. "We also note that this new focus by and competition between Juniper and Cisco will likely impact other, smaller players in the market."

The Layers 4-7 switching, or traffic management market, is similar to WAN optimization in that it focuses on the efficient use of bandwidth consumed by applications. Big players in the market include public companies such as Radware Ltd. (Nasdaq: RDWR) and F5 Networks Inc. (Nasdaq: FFIV), as well as startup NetScaler Inc. and even U.K. newcomer Zeus Technology.

”It has brought the financial analyst community to our front door,” says John Oh, director of corporate communications at NetScaler. “When a public company the size of Juniper makes a move it not just validates the market but it typically causes a domino effect.”

Jon Rabinowitz, senior marketing and communications manager at Radware noticed a lot more potential customers coming to his company’s booth at this year’s Interop. “A lot of people are asking questions,” he says. “It has raised the profile of this market once again.”

This is hardly surprising. During the Interop show, Juniper’s CEO, Scott Kriens, identified the potential market for traffic management products as being worth “billions and billions” of dollars over the next few years (see Zeus: A Takeover Target? ).

”There’s a buzz around this space,” says Jeff Barker, director of product management at Packeteer Inc. (Nasdaq: PKTR). “People coming to the booth are not asking ‘What does this do?' They are asking ‘How do I deploy it?’ "

— R. Scott Raynovich, US Editor, Light Reading, and James Rogers, Site Editor, Next-Gen Data Center Forum

Page 1 / 2   >   >>
materialgirl 12/5/2012 | 3:15:55 AM
re: Cisco Prowling WAN Optimization? I am shocked! Shocked! that IOS cannot handle this processing load. What about the new OS on the CRS? Perhaps ported to a 12000 it could handle a 100Mb link with some level of inspection?
jsittko 12/5/2012 | 3:15:53 AM
re: Cisco Prowling WAN Optimization? The path of nearly all hardware companies, when faced with margin compression or perceived need for more added-value, is services. The problem often discovered was thay they did not really understand this business and they tended to be "body shops."

CSCO could layer on a built out optimized network immediately.

Admittedly, a wrinkle would be that CSCO would now be selling hardware to companies that are now competitors to some degree...so how much hardware would they continue to sell?

But think about it. When seeking value-add companies have either migrated out of the core towards access or moved up the stack from phy to application.

It's a path I see as inevitable.

dwdm 12/5/2012 | 3:15:49 AM
re: Cisco Prowling WAN Optimization? I have tested OER in a semi production environment. It does work fairly well in a WAN border router (without using an external controller), but for sure it is not for an ISP. OER manages network routes and paths based on information from other features. So the more paths and prefixes you got the more load on the CPU. My tests included routers with about 4 exit paths and about 3000 routes with no issues. A friend of mine that works at another company is also testing it without any issues either (although he is testing it over tunnel interfaces). For what it is designed for, it works very well. And it has nothing to do with what Peribit does.
donniall 12/5/2012 | 3:15:47 AM
re: Cisco Prowling WAN Optimization? So I thought the Corvil capabilities that Cisco is (supposedly) planning to 'bundle' with router platforms is their WAN optimisation offering ???
Don-
dwdm 12/5/2012 | 3:15:45 AM
re: Cisco Prowling WAN Optimization? Don't know much about Corvil. I looked it up in Cisco's docs, and it looks pretty interesting... Just came out in the latest 12.3T release.

It seems that it is a tool to estimate the required bandwidth for certain applications (that you match with an ACL). So the tool tells you how much bandwidth is required for an app, and then you're supposed to tweak your QoS accordingly. Sounds pretty cool... but I haven't used it. Setting QoS parameters has always been black magic, but with something like this I can see some nice value.

What Cisco is lacking is decent compression and TCP optimization. This is the value of Peribit. Although I have no idea how they perform from an aggregation prespective.
SS7 12/5/2012 | 3:15:44 AM
re: Cisco Prowling WAN Optimization?

If Corvil is depending on IOS for delivery of its performance management.....good luck.

Customers, especial big ones w cash are very sensitive to the IOS level. Ever hear of Safe harbor?

Getting fancy reporting from IOS is not enough to have them upgrade to specific IOSGǪtoo much risk involved.
ron202 12/5/2012 | 3:15:43 AM
re: Cisco Prowling WAN Optimization? I taught that for a good compression you need a large amount of data so so the latency will increase -not good for voice. Plus some of the files (jpeg , mpeg , mp3 etc) are already compressed.
Does Peribit has new mathematical patents in the area of compression?
dwdm 12/5/2012 | 3:15:41 AM
re: Cisco Prowling WAN Optimization? I agree with SS7. Nobody in their right mind would deploy the latest 12.3T in production in any decent size network. Although the severity of IOS issues in new releases has declined in the recent years, it is still not safe. But they did deliver the feature, so I would assume at some time later this year, there will be a more "baked" release out there with such functionality. It always works this way.. nothing new. Same applies for any other company.

By the way, I did hear about the safe harbor. We used to follow the GD status for IOS, but it is worthless. The safe harbor is great. Safe harbor is a very positive move for Cisco.
donniall 12/5/2012 | 3:15:38 AM
re: Cisco Prowling WAN Optimization? The renewed growing emphasis on WAN optimization technologies is an interesting trend.... In the late 1990's the general feeling was that bandwidth costs were falling dramatically, and rather than deploy technologies to 'conserve' bandwidth (as a typical scarce resource) it was considered more prudent to deploy technologies that would use the available bandwidth more efficiently (robust QoS etc) ...Now it seems we have come full circle ..? The focus once again appears to be on conserving bandwidth as a scarce resource ..This co-incides with another significant trend: equipment vendors now openly advocating more intelligence within the network - thus compromising the e2e principle of the internet (intelligent end-points, dumb netowrk). This trend is even more apparent in the ITU driven IMS initiative (and Junipers Infranet initiative) ..... Its also interesting that Cisco now openly talk of the 'intelligent network' (in fact Chambers even used the expression himself talking about yesterdays earnings announcement) ..
Any perspective on what these trends indicate? Are we destined to see the old business models prevail with large service providers being effective gatekeepers to what can/cannot traverse the network, and meanwhile bandwidth retail prices being kept artificially high ....
A worrying development if that is the case ...
Don-
lite_network 12/5/2012 | 3:15:38 AM
re: Cisco Prowling WAN Optimization? This is an interesting observation. I have also noticed that Cisco and Juniper are trying to put as much "intelligence" (Application knowledge) in the network as possible. This seems to be a response to the Chinese getting into the routing market (Huawei, ZTE).

The Chinese have been able to make quite competent Routers (Huawei, NE40e, NE80e) which follow the current standards quite well. This strategy seems similar to the PC vendors. When hardware started to becomme a commodity, software (microsoft) became the focus and the place for the most value.

One small issue with this strategy is that it depends on keeping this application knowledge proprietary (like Microsoft). There are currently little or no standards on how to recongnize flows, application sessions, security threats. I expect that soon the buyers of this equipment will demand standardizatio and that is where trouble will start.

-LN
Page 1 / 2   >   >>
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE