Cisco this week is set to roll out a suite of products that will help network managers better pinpoint the source of application performance problems.
Cisco's Network Application Performance Analysis (NAPA) launch delivers on the company's promise to develop more technology to manage network and application performance.
Using technology acquired through its purchase of Sheer Networks, OEM licenses with Opnet and Corvil, and in-house technologies, Cisco built four products that the company says will help customers identify application performance issues. The tools will also make it easier to pinpoint the cause of the application slowdown in the network. The four products work independently of each other, but will also be more tightly integrated to work together going forward, Cisco says.
The products set to be available this month include Application Assurance Solution (AAS), which is software that uses distributed agent technology to capture application packets, network metrics and traffic data to perform deep dive analysis on specific network and/or application performance problems.
Also new is Network Planning Solution (NPS), another software product that collects data from network components to provide a virtual view of the network to perform capacity planning and determine network readiness for new application rollouts, such as VoIP.
Also available this month is Bandwidth Quality Appliance (BQA), which for the moment is a product coupled with Cisco Advanced Services. Cisco says it will make the appliance available without requiring services in future releases. The appliance installs near WAN links to provide granular data regarding bandwidth use and application traffic over the wide area.
Set to be put into beta tests in January is Performance Visibility Manager (PVM). This software product installs on a server and collects data via NetFlow, SNMP and other protocols to provide a high-level view of network and application performance. Cisco officials expect the software to be generally available in March.
"Cisco is sending a clear message to the market that they are very serious about getting into network and application performance management," says George Hamilton, a senior analyst with Yankee Group. "This is a direct shot at the HP OpenView, IBM TEC [Tivoli Enterprise Console] and other traditional management tools out there. Cisco put a lot resources behind this internally."
Cisco products that alert network managers on application performance could in the short-term help performance-management software makers such as InfoVista, NetQoS, NetScout, which today go to great lengths to incorporate knowledge of Cisco gear into their software. With Cisco providing its own standards-based management tools, it will be easier for such vendors to manage Cisco nets without having to overhaul their software, Hamilton says.
But the push is not without its challenges. Cisco will have to maintain its gear to keep current customers happy. Perhaps more difficult, Cisco may have to sell to systems administrators, as application performance in many IT shops is directly linked to specific servers.
"Cisco is a huge networking equipment vendor, and trying to change that mindset in the market will be tough to do," Hamilton says. "Cisco wants to position the network as the platform for data center automation going forward."
In fact, Cisco believes the network is the where companies should be rolling out virtualization.
According to Clive Foreman, a vice president in Cisco's Network Management Technology Group, the company will in 2006 announce network virtualization tools, currently under development with the help of technology acquired through its purchase of Topspin, a maker of grid computing technology.
As Foreman explains, when other products are used to provision or reposition servers to accommodate load on demand, there is also usually a network component that need to be reconfigured to support the system change. As companies such as IBM and HP, virtualize the storage and server layers, Cisco will step up to virtualize the network layer and enable more automation when such changes occur.
"Today customers probably have a great server tool, but someone still had to go and manually update the network settings," Foreman says.
Yankee Group's Hamilton explains that server and storage virtualization is done on more component-by-component basis, but network virtualization can touch all infrastructure components and represent the resources among separate technology pools, for example, server, storage and application.
"The network is the consistent component across the entire infrastructure, and it can make more sense to virtualize that than support virtualization across isolated pools of technology," Hamilton says.