Sometimes it's not immediately obvious why two companies might want to combine forces, such as in the case of F5 Networks buying Acopia Networks. At first glance, some might wonder why F5 would be interested in a storage company.
But Acopia is not a storage company, says former Acopia CEO Chris Lynch, who today is senior vice president of F5’s data solutions division. I had a chance to talk to Lynch recently about the deal with F5 (which was finalized in September), how the two companies’ technologies complement each other, and opportunities for future product integrations.
Why did the union of F5 and Acopia make sense?Acopia is Spanish for “to gather together,” and the idea behind the company was to inject a new level of intelligence into the fabric of the network that would allow you to control storage and servers, and hence, commoditize them. We’re really an amalgamation of systems thinking, network thinking and storage thinking, brought together. We went off and solved this problem of file-level virtualization and data management. Ninety percent of applications in the enterprise address files. So the fact that F5 evolved from the Web server layer into a broader applications orientation, and we moved below them to the data layer, was an interesting synergy. We’re complementary from the standpoint that F5 optimizes the application layer, and Acopia optimizes the data below that layer. From an architectural standpoint, as we started to talk, we saw that we both leverage intelligent proxies to gather the smarts that we need, and the techniques that we use to inject that into the network are similar. So from a technical standpoint, there were all these synergies. We also had a lot of cultural synergies, as companies.Did both companies also target the same person within the IT organization, from a sales standpoint?Not 100%. But in F5’s most successful deployments, they’ve gotten mindshare from the CIO and from people who are tasked with owning applications. Their most successful, broadest implementations have that sponsorship, as do Acopia’s. If it gets pushed down to a more tactical level, F5 typically ends up with a network person, while Acopia would end up with maybe a storage administrator. That can work, but it’s not optimal.Is there any thought of combining the features of F5’s application delivery appliances and Acopia’s file virtualization switch into a single device?That’s a question that a lot of customers are asking. The simple answer is that there isn’t a restriction in terms of where we sit. We’re actually a pure network switch, we’re a full-blown 3.2Gbps switch, and we connect into the Layer 3 infrastructure. We connect almost like a one-armed device into the Cisco Catalyst infrastructure. We sit at the right layer to do it, even though what we’re really controlling is the back-end storage. So there’s not a limitation in terms of the real estate. We’re actually sitting in a place were you could combine the functionality. Today we have not announced a plan to do that. But I think it’s an astute observation that if application and storage virtualization are converging, why wouldn’t you provide all those services in one multipurpose device? That is a vision point.What have you been working on so far, in terms of integrating the two product lines?We have in the first 90 days worked on integrating into F5’s enterprise management system, which is a big improvement over the one that we came into the company with. We also are working F5’s DNS intelligence into the switch, so that we now have the additional intelligence to understand geographic proximity of clients and data, which we get from F5’s DNS server technology.For big-picture projects, like data center consolidation, is that where the two products really work well together?That’s probably the single biggest one, data center consolidation. Also, rollout of Web-based applications is another disruptive event that creates an opportunity for the combined entities.Can you share some details about product direction for 2008?WANJet product, which is now under my responsibility in the data solutions division. There are a number of synergies that we’re going to apply. Think about taking the Acopia global namespace and tying that to the WANJet intelligent proxies and WAN acceleration devices. We’re combining those things so that we basically can provide a policy-based distribution of data that will be centrally controlled, secured and managed. Right now among all of the technologies out there -- like the Riverbed and Cisco products -- none of them have the concept of a global namespace. We’re combining WAN acceleration with virtualization so that back at the data center, when you’re backing up data centrally, for instance, you’re not taking down servers or storage or applications to do so. That’s a key distinction.
I think you will see in 2008 significant integration with the capabilities of F5’s
Any final thoughts?Acopia gets associated with storage because the benefit is so dramatic. Our customers typically save 50% on the capital and operational expenses related to storage when they deploy our technology. So because of that, people view us as a storage company. That was one of the reactions: Why did F5 buy a storage company? But we’re really not, we’re an intelligent network switch that optimizes storage. That’s one of the key messages. We’re a network IP-based switch that operates the data management of unstructured data. That results in huge storage savings, but it doesn’t make us a storage device.