News

Cisco boosts WAAS

Andrew Hickey

Cisco this week released key updates to its Wide Area Application Services (WAAS), partnering with NetQoS for better visibility into how applications are performing over the WAN.

At Cisco Networkers, the networking giant's annual user conference, Cisco announced a partnership with NetQoS to provide end-to-end visibility and application monitoring of applications traversing the WAN. Adding metrics from NetQoS illustrates application response time, application data rates, link utilisation and protocol breakdowns before and after optimisation so networking and applications pros can see how traffic is performing on the WAN.

According to Steve Fulton, senior director of strategic alliances for NetQoS, companies are struggling to get the most out of their WAN optimisation deployments but are lacking a way to measure applications' end-to-end performance. Most WAN optimisation tools measure TCP application response time and other metrics in segments -- the client, WAN and server segments -- which can be misleading as to what true response time is. The client segment shows the transaction time, or end-user experience, while the WAN segment shows the network roundtrip time, or latency. The server segment shows server response time and back-tier performance.

Fulton said the Cisco partnership will integrate NetQoS monitoring capabilities into every WAAS device, the Wide Area Application Engines (WAE) and Wide Area Application Engine Network Modules (NM-WAE)

    Requires Free Membership to View

to look at application information before and after it is optimized by WAAS.

The integrated software on Cisco WAAS devices will export TCP header information before optimisation occurs to NetQoS SuperAgent, allowingWAAS and NetQoS customers to quantify response time improvements, improve troubleshooting, and deliver constant and improved application delivery.

To measure traffic flow improvements from Cisco WAAS, the two vendors will also market the NetQoS ReporterAnalyzer, a traffic analysis module that reports on Cisco IOS NetFlow statistics. According to Cisco, WAAS provides compliance with network-based NetFlow collection and export and does not impede the ability to see traffic composition before and after WAN optimisation is applied.

Dr. Issy Ben-Shaul, CTO of Cisco's Application Delivery Business Unit, said: "Cisco WAAS integrates transparently with customers' networks, and the management interface we have developed with NetQoS extends that transparency to performance reporting data, giving our customers accurate insight into how to best leverage our technology to maximize application performance and remote user productivity."

The Cisco-NetQoS pairing illustrates and validates WAN optimisation results. It also lets users identify and prioritize the network segments and applications that will benefit from WAN optimisation; obtain before-and-after measurements for WAN optimisation pilot testing; determine optimisation's impact on application performance and data volume across WAN segments and in the data center; troubleshoot problems after deploying WAN optimisation; and calculate the ROI of WAN optimisation deployments.

"Without this, customers were using a stopwatch for their [WAN optimisation] rollout or relying on somewhat inaccurate information," Fulton said. "It was a band-aid approach to a bullet wound."

Burton Group senior analyst Eric Siegel said adding monitoring capabilities into WAAS allows WAN optimisation problems to be detected immediately, before users find out, creating end-to-end visibility. He said simply measuring response inside just the data center, instead of end-to-end, offered very little ability to resolve issues.

In addition, Siegel said, identifying optimisation problems early allows for a sort of triage, where responsibility for a problem can be determined.

"With this kind of measurement, you can see if it's the network, the server or the application," he said. NetQoS adds that capability into the WAAS device, to identify whether the issue is between the box and the server or between the box and the user, or whether it's the WAN connecting the boxes.

"It's good to know really quickly what's going on, so you can know if you're really screwed; otherwise it's a couple of hours before you know for sure," Siegel said.