WAN optimization: Tactical or strategic or both?
- 25 July, 2008 10:20
As you may have suspected from some of my prior columns, I've found the product strategies in the WAN optimization market something of a puzzle.
IT management is an industry that is ambitiously skating towards a revolution -- toward a more Lego-like world of interconnectedness (not to sound too much like someone from the Sixties). In contrast, briefings with many of the WAN optimization crowd reminded me of my days supporting IBM marketing for Token Ring - all speeds and feeds with little interest in end-to-end visibility and strategic integration. The fact that this is beginning to change is a good thing, in my view.
WAN optimization products typically provide WAN data services that accelerate application performance. WAN optimization technologies are varied and may include application compression, packet payload optimization, and Wide Area File Services (WAFS), among others. These are all focused on individual links and most commonly deployed at critical junctures with remote branch offices.
The issues have been that while WAN optimization products deliver real value on a link-by-link basis, they obscure end-to-end visibility so that strategic planners are left to guess at the real "before" and "after." They are also typically housed as an appliance with, up until very recently, little interest in integration with the rest of the management community. Finally, as "cheese-stand-alone" solutions, they become costly when scaled to large enterprise environments with hundreds and sometimes thousands of remote locations.
This somewhat ambivalent position was underscored in some EMA research just completed across 180 respondents primarily in North America and targeting a whole array of issues surrounding the changing role of networking and network operations. These ranged from a focused set of questions on managing Web applications over the network, to responsibilities in managing non-networking devices in remote locations, to interest in UC and emerging technologies, to CMDB System perceptions among networking professionals.
The questions on WAN optimization fit into the broader set of managing applications over a network and produced the following rather striking results. In the survey, when asked if WAN optimization was tactical or strategic, tactical won out 46 per cent to 44 per cent, with the rest abstaining. The most dominant reason for being strategic was "They support my strategies for virtualization across the infrastructure." This in itself gave me pause. I parsed this to mean "They support the strategies I wish I had for virtualization across the infrastructure." The next two were tied and close -- saving in bandwidth costs and allowing for the assimilation of remote locations more effectively. These have been the mainstay of the WAN optimization market to date.
The slight majority's number one reason for viewing WAN optimization as tactical was its poor capabilities for integration, plus "They will be replaced by newer technologies that provide better end-to-end path optimization in the future." I won't comment on this except to feel a little bit vindicated after some of the conversations I've had in the past on integration and product design. The third reason to vote for tactical was "They are too expensive to scale to my infrastructure."
The good news that in the past year and especially in the last six months, the industry has seen some moves to break out of its silo. Most notably, in recent months, Riverbed has integrated with NetScout's nGenius Performance Management System and then announced intentions to integrate with OPNET's products. Prior to that, Riverbed announced integrations with Mazu Networks, which has in turn also integrated with Packeteer/Blue Coat. Last summer, NetQoS announced integration with Cisco's Wide Area Application Services (WAAS). And other vendors in WAN optimization have also made efforts to step into the broader market - e.g. Expand, through its virtualization initiative, and Ipanema with its service impact directions across the WAN.
To highlight the benefit of one case in point, let's look at the NetScout/Riverbed integration. NetScout's nGenius Performance Management System has been granted enhanced visibility into Riverbed's proprietary signaling protocol in its Steelhead appliances. NetScout's broader view of the network across many multiple links and its analytics and reporting can then be brought to bear in establishing a clear "before" and "after" context for planning and assessing the value and effectiveness of Riverbed deployments.
This can be used for planning and optimization ("where will Riverbed be used most effectively?") and can also support planning for new applications. It also brings a capability for troubleshooting and fine-tuning existing Riverbed implementations and application performance issues where a Steelhead appliance is deployed.
I have to say that I am relieved to see this trend, and I expect it to continue and to, no pun intended, accelerate. Stepping up to the 10,000-foot level, there are two overall trends in the management market that are defining present-day and near-term advances - more cohesive visibility into services and their interdependencies; and more effective approaches to automate actions, whether machine to machine, people to machine, or people to people in workflow. These combinations to bring visibility with automated control in WAN optimization are now, finally, right in line with mainstream requirements and mainstream demand.