Showing posts with label Riverbed. Show all posts
Showing posts with label Riverbed. Show all posts

Wednesday, February 20, 2008

Riverbed CEO: 'We Can Be One of the Survivors'










Jerry Kennelly, chief executive and co-founder of networking equipment maker Riverbed Networks (RVBD) stopped by the Barron’s offices today after three days of meetings with investors. He offered two takeaways: the company is not for sale, and making his sales forecast this year shouldn’t be hard.

Given that Riverbed shares spiked 9% yesterday, as Eric noted, I asked if RiverBed was for sale, to which Kennelly said, “None of use wants to work for Cisco [Systems (CSCO)] or one of the giants,” portraying Riverbed’s staff as a a tight-knit band of humble software engineers spending all day on their laptops at Starbucks (SBUX). Not exactly the Cisco culture, in other words. Kennelley argues Riverbed “can be one of the real big survivors in tech over time” and so he’d like to go the distance.

He attributes the bounce in the stock to a couple things he’s talked with investors about this week. First, he sees it as no big deal for the company to make sales of $375 million this year, as the company implied last week. As he sees it, “In the last quarter of 2006, we were at a $140 million annual revenue run rate, but we ended up delivering $236 million [in sales] in 2007. So, if we were at a revenue run rate of $300 million in this most recent quarter, we think it’s reasonable to expect we can deliver $375 million this year.” RiverBed did $76 million in the fourth quarter ended in December. And the uses for Riverbed’s gear are newer than for Cisco’s, which means, he thinks, CIOs will cut back on Cisco’s equipment before they cut back on Riverbed’s.

Second, the company has plans to spread its technology further as the year goes on. Today, Riverbed gear is used mainly for speeding up how fast branch offices can fetch data from computers at headquarters. New versions of the company’s switch could improve how efficiently data is moved between central computers within and between corporate data centers, conceivably broadening Riverbed’s addressable market. Kennelly is holding off on details of the product until its formal introduction.

Friday, January 25, 2008

Will Cisco ever catch Riverbed or should they just get out now?

Source: CiscoSubnet

I want to start off with an excerpt from a Gartner paper by Mark Fabbi, it was put out back on March 15, 2007. This paper was written for the Cisco Applications Network Services, an before we go farther the Cisco WAAS is a part of this service. Plus we can look at it as a standalone product.

From a product perspective, Cisco's progress through 2006 has been, at best, mixed and largely disappointing. Cisco is a late entrant to an innovative and dynamic market, and it is "playing catch-up" across all market segments. Despite new product announcements during 2006, Cisco still lags behind the market.

I answer so many questions each week from Cisco customer with regards to the Cisco WAAS, what they are told by their Cisco Sales teams and the Riverbed Steelhead. Being a partner of both products it puts me in a unique position with my customers, it allows to me to what is right for my customer. Many VAR’s do not have this opportunity and are looking just to sell a customer more equipment. I always remember the excerpt above that was put out by Gartner and keep waiting for a change, but I have been disappointed just like customers.

From a VAR stand point I think it is time for Cisco to cut its losses and put the WAAS on the shelf. Right now it is not even in the top two of the product category, Riverbed Technologies and Juniper are one and two. If you are asking why would I say this? Well I have been involved in many bake off’s with customers using every product in this category, when it is an unbiased test Riverbed has never lost. By unbiased I mean that that the VAR just does not drop off equipment to the customer and says go test. You cannot just do that with customers or customers will pick based on culture or who gives them more free stuff than who really is the best.

Case in point, I will give two examples of what has happened to me when I was involved in two evaluations using Cisco, Riverbed (product demo) and Juniper. The first was a state organization that would not allow our engineers to be present when they did the testing; they just wanted the equipment and us out of the way. They said that if we were there we would influence the decision and that they were a state organization, let me say the biggest in the state. Now I have performed many bake offs and with the applications they were using, Riverbed would have no trouble winning.

It came to be that the customer said that Juniper and Cisco both did better than Riverbed, but they would not show us the data. Another bake off was with a very large company in which they were drinking the Cisco Kool-Aid way to much and an engineer sabotaged the evals so that Cisco would win. . If they were so up on putting the wrong product in, you might was well let them and we did. Some people will only put Cisco in and they do not care about anything else. There is an old rule with IT Managers and Directors; you never get fired for putting Cisco in your network even if it does not work. This same company called me one month later wanting help with the WAAS when they could not get it to scale or work like they did in the testing.

So as you can see, even if you are number one with Gartner you still have a huge mountain to climb when going up against Cisco. Now back to another reason why Cisco needs to drop the WAAS, the new mobile client from Riverbed is leaps and bounds above anything anyone else has in the network space. Cisco is years away to catching up with Riverbed on the WAAS and another three to five on to that just to get a mobile client that could compete with Riverbed.

I have copied the cautions from the 2007 Gartner report on WAN Optimizations Controllers; both of these were with reference to the Cisco WAAS:

* Successful implementation often requires multiple days by on-site Cisco engineers, dueto the solution's complexity. There is no single view of configuration, policy or WANoptimization features that are split across Wide Area Application Services (WAAS) andseveral router-based Internetwork Operating System (IOS) software options.

* Cisco has been slow to understand emerging market needs, resulting in WAAS featurereleases usually following other vendors' innovations. For instance, WAAS lacksadvanced features such as acceleration for HTTPS and Messaging ApplicationProgramming Interface (MAPI), and Cisco does not offer a software WOC client. Each ofthese features is available from other vendors.


Now I know things change and Cisco has fixed some issues, but even in this December 2007 Gartner Report they did not even mention the lack of a mobile client from Cisco. Now just to be fair I am going to show what Gartner put in about Riverbed:

* Less-capable QOS and reporting features than some leading vendors.

* Steelhead lacks UDP support.


Now I have to say that it looks to me like Gartner could not find anything else to write about Riverbed since they have great reports and have no issues with QOS (Quality Of Service)for any customer using a Cisco or Juniper network.

Let’s think about this for a second, what would happen if Riverbed took my ideas of which I have told the CTO and published a Windows Mobile Client, Blackberry Client and then got into the home user market?

This is my unbiased opinion being a partner of both vendors and doing many bake off’s. I hate to say this about any Cisco product or any product one of my vendors have but if you want to bring order to chaos within your network, don’t use the WAAS. Going back to my comment above it is about doing what is right for your customer with any product; give them the best product for the money.

Time to get out Cisco, maybe a Cisco Iphone?

IF YOU NEED TO PURCHASE HARDWARE FROM RIVERBED PLEASE CONTACT ME BY PHONE +31650730710 OR BY MAIL DESIGNFORIT@LIVE.NL

Sunday, January 13, 2008

WAN optimization: Money for nothing, clicks for free

Application Acceleration and WAN Traffic Optimization : Market Trends

Optimization technology can slash the cost of WAN bandwidth and improve app performance

Source: Network World

WAN optimization technology is a way to save money — take that to the bank. For instance, custom machinery manufacturer Curt G. Joa in Sheboygan Falls, Wis., avoided the cost of adding a server to its German office by installing WAN-acceleration gear from Silver Peak.

Joa engineers in Germany needed to access engineering applications, but the applications performed so slowly over the WAN that it could take an hour just to open a Novell directory.

The initial plan was to install a server in the German office for $100,000 including time, hardware and licenses. Instead, the company spent $17,000 on a Silver Peak NX appliance that solved the problem by speeding application performance over the WAN.

In another example, Riverbed gear that cost $20,000 reduced the need for bandwidth between a Millard Lumber site in Omaha and a second site in Lincoln, Neb., from three T-1s to one, saving $3,000 per month. That’s a payoff within seven months.

In addition, WAN-optimization devices improve transaction times between sites by as much as 90% or more, reducing user frustration and in some cases making it possible for applications to perform at all over long distances.

Despite their impressive results, these devices often are overlooked, mainly because potential users worry they might remove chunks of data, block visibility of traffic through firewalls, or fail and stall out networks altogether, says Eric Siegel, an analyst with the Burton Group.

These concerns are groundless, Siegel says. “They should do it now. You can really save money on these things. And there’s soft benefits like happier customers and more cooperation,” he says.

In the grand scheme of things, a paltry amount of WAN optimization gear has been sold, says Mattias Machowinski, an analyst with Infonetics. This year, vendors will sell about $300 million worth worldwide.

Even so, the optimization gear is catching on. Sales this year have increased 27% compared with last year, and Machowinski projects double-digit increases in each of the next three years.

These devices help save money in two ways. First, they reduce the need for WAN bandwidth, which translates into buying smaller WAN pipes or staving off the need to add more.

Second, they enable businesses to consolidate servers in data centers, saving money by reducing the number of servers in corporate networks that need to be purchased, installed, maintained and repaired.

Optimization appliances use a variety of means — compression, caching, boosting TCP efficiency, protocol optimization, imposing QoS — to reduce the amount of traffic that crosses WAN circuits, compress the traffic that is sent and make sure it does so efficiently to avoid congestion.

Vendors — Cisco, Citrix, Expand Networks, F5 Networks, Juniper, Packeteer, Riverbed and others — use varying blends of WAN-optimizing technology, so one vendor’s gear might do a better job on a particular traffic mix than another’s. Consequently, the single most important thing customers can do is test gear made by more than one vendor on live networks.

IF YOU NEED TO PURCHASE HARDWARE FROM CISCO, JUNIPER, RIVERBED, PLEASE CONTACT ME BY PHONE +31650730710 OR BY MAIL DESIGNFORIT@LIVE.NL

Friday, December 21, 2007

Engineering firm saves $1 M annually with network efficiency appliance



Source: Computerworld

With 850 employees working out of 20 offices scattered across five Western states plus Mexico, Psomas is a large civil engineering firm founded in 1946 and named for its founder, George Psomas. It focuses on projects that cover the spectrum of civil engineering, from water/wastewater infrastructure to highway designs to land development engineering. Since 2005 it has saved more than $1 million a year in network costs by implementing the Steelhead network technology to improve the efficiency of its WAN transfers of large CAD files between offices.

As Psomas has grown over the years, it has sufferred increasingly from the simple-seeming problem of getting the right talents on the right projects. Psomas has always believed in work sharing among offices, partly to balance workloads and partly to match the people with the right skills to the right projects. The "home office" for a project usually is the office closest in physical distance from it, but some of the people with special skills that a project requires may live and work hundreds of miles away, in another state or across the U.S./Mexican boundary.

What Psomas found itself doing was transferring large CAD files over the network between offices overnight. These files are huge, and often one CAD file references several others, creating major network loads. This brute force, manual approach also created problems with version management, sometimes resulting in the overwriting of files, requiring engineers to redo work.

"We knew we had a problem," says Chris Pinckney, Psomas's CIO. "But we didn't know there was such a great solution out there until we ran across Riverbed at a conference." They had looked at other potential solutions, but none of them fit the company's needs or fit with its working methods.



Riverbed provides an enterprise-class, appliance-based solution for accelerating the transfer of large files over corporate and public WANs by several times -- on its Web site the company claims up to 100X acceleration increases. The technology is based on Riverhead's Steelhead appliances and Steelhead mobile software. After evaluating the technology, Psomas finished its roll-out of 15 Steelhead appliances to its larger offices in January 2005. It uses the mobile software version for remote employees and within its smaller offices, which do not have the volulme of file transfers to justify an appliance. Since that rollout, the technology has had a definite positive impact, basically solving Psomas' CAD file transfer problem even as its business, and the volume of workshare projects, has grown.

In its initial purchase proposal to senior management, Psomas IT justified the cost on the basis of increased productivity due to increased workshare and decreased need to redo work lost in accidental file over-writes. "And we have seen that," Pinckney says. "But we ended up cost justifying it on bandwidth." These savings come directly from avoiding line upgrades. Monthly data traffic reports typically show a 3X reduction between the files going through the internal LAN and the traffic across the WAN on the other side of the Steelhead appliance. This has allowed it to save a net of more than $1 million a year in direct network costs.

For instance, in October 2007 Psomas saw 2.4 Tbytes of data on its LANs optimized down to 823 GB of WAN traffic by the Steelhead technology, This is an operational figure, not an optimized test, making it all the more impressive. The reductions are so dramatic that they raised questions. However, Pickney says, at the most recent Riverbed users conference, the CIO of another large organization reported that they did independent measurements on both sides of their Steelhead appliances and found the reporting accurate.

This increase in WAN efficiency has allowed subsidiary savings. For instance, Psomas has been able to reduce from seven Exchange servers to three and to centralize most of its mission-critical IT operations in its Irvine, Calif., collocation facility. That facility has required increased trunk lines, but that has been more than made up for by the efficiencies of centralization and reduction in trunk line demands at its other offices.

This approach to ROI, however, is one reason its smallest offices, typically 10-15 employees each, do not have Steelhead appliances. They are not always engaged in workshare projects and so do not have the volume of CAD file transfers that the larger offices generate.

"That is where the Steelhead mobile product comes into our picture," Pinckney says. "It does the same magic the appliances do" and is sufficient for offices where only a few engineers would be collaboratingwith projects in the larger locations at a given time.

Riverbed's capabilities have long-range implications for Psomas as well. "We are seeing a real drop-off in the number of new civil engineers graduating in the United States," Pinckney says. For instance, one of the major California engineering colleges only graduated 50 civil engineers last year. "I have heard experts say that in the future we will be depending more on civil engineers from Asia." At the same time, the demand for civil engineers has increased. They are the master builders of everything from skyscrapers and bridges to water treatment and groundwater facilities. Therefore, an engineering firm like Psomas needs to get maximum efficient use from the engineers it can recruit, and part of that is using the network cloud to connect experts, wherever they may be located, with projects often in geographies far from the location of some of those experts.

Overall, Pinckney says, "When Psomas found Riverbed, it was a hallelujah moment."