SHARE
Follow this article on Twitter Facebook LinkedIn Bookmark and Share
Home >> Communications Infrastructure

Riverbed Cascade 8.4 measures network performance

Riverbed Cascade 8.4 measures network performance

By:  Greg Meckbach  On: 27 Jan 2010 For: Network World Canada Creator

With the new release of its network performance management software, Riverbed Technology Inc. said administrators can find out how long applications take to respond to individual users. It works with the Steelhead WAN optimization appliance. Yankee Group's Zeus Kerravala says virtualization could help drive demand

Riverbed Technology Inc. (NASDAQ:RVBD) has announced version 8.4 of its Cascade product, designed to measure network performance from the user’s point of view.

The software is designed to monitor the performance of applications and networks. It works with Riverbed’s Steelhead hardware, which is classed as WAN optimization because it is designed to improve performance of applications running between branch offices and central data centres.

“If you are putting acceleration on the network, acceleration changes a lot of the behaviour of the network,” said Yoav Eilat, Riverbed’s director of product marketing for Cascade. “Here we try and tackle the problem of giving full reporting. We want to see if they are getting the performance they are usually getting or if someone is slowing them down.”

San Francisco-based Riverbed said Cascade provides information on servers, applications and how they depend on each other. Version 8.4 shows the rate at which data from an application is traversing the network, taking into account retransmits and network overhead. It also measures the duration of connections for transaction processing applications and the throughput users experience while their connection is actively transmitting.

It costs US$495 per site.

“Every user that uses an application will do some work for a while, will be active and then inactive,” Eilat said. For example, a salesperson using enterprise resource planning to create sales orders might read the screen for a period of time, or do something else offline, before their application connects to the network.

“The key is to pinpoint when a user is active and see how fast they are able to go,” he said. “There’s a bunch of different metrics that we’re measuring.”

He added the aim is for the administrator to detect problems that would prompt users to call the help desk.

Identifying problems has always been an issue in WAN optimization, said Zeus Kerravala, senior vice-president at  Boston-based Yankee Group Research Inc.

“Nobody else really has a top down view of how the response time appears to the users,” Kerravala said.

Eilat said if performance drops and it’s related to WAN use, a link may be overloaded and the administrator wants to see what causes it. Using Cascade 8.4, he said, an administrator would drill down and find out which application is affected, drill down again and find out which server is affected. Once you determine a server, he said, you could narrow it down to determine which individual user is responsible for a spike in traffic.

“Whichever way you want to slice and dice it, you can see how the performance is,” Eilat said. “You can see problems users would end up calling you about.”

As more companies use virtualization to combine different services on to one server, administrators are having a harder time measuring network performance, Kerravala said.


Sign up for our Newsletters












Print |  Views: 4724   |   Rating:offoffoffoffoff  (0 votes)
Rate this article on a scale of
1 to 5 stars,5 being the best.




Greg Meckbach Greg Meckbach Greg Meckbach is editor of Network World Canada and has worked for ComputerWorld Canada, Communications & Networking and Computing Canada.

Comments (0)

No Comments!
Name: (required) eMail: (optional)

Your email address will not appear online and will be used only if the editor wishes to contact you personally for additional comments.