Apparent Eyes Proactive Network Fixes

Beta testers say the AppCritical network testing tool from Apparent Networks has real promise.

Apparent Networks is the latest in a long line of network and systems management vendors talking a good game about how they can help administrators manage proactively.

The difference this time? Apparent Networks may actually be able to deliver, said Jimmy Brown, vice president of network services for Dallas-based outsourcer Affiliated Computer Services.

Brown is beta testing new network management software from Apparent Networks that can test the different paths that applications traffic traverses and automatically identify problems by the signatures they create as packets cross the network. The offering, AppCritical, will be launched May 1 at Interop in Las Vegas.

Brown said he has found that the new AppCritical network performance management system can find problems and allow ACS and its clients to fix them, often before users become aware there is a problem.

"Proactive monitoring is what AppCritical delivers, so we can ensure that we understand whats going on with the network at all times instead of just reacting to client requests," Brown said.

The patented technology in AppCritical allows the tool to perform continuous, real-time testing of network paths, rather than focusing on the availability of individual devices in the network. It sends out a small burst of precisely timed and configured test packets once per minute, but users can define other interval periods.

"When we detect somethings changed ... we can then test more frequently," said Kelly Daniels, chief technology officer at Apparent Networks, in Vancouver, British Columbia.

As it detects problems, AppCritical automatically escalates the testing, performing deeper and deeper tests to look for such Sublayer 3 problems as full/half duplex mismatches, NIC (network interface card) driver problems, rate-limiting behavior, QOS (quality-of-service) issues and so on.

/zimages/3/84833.gifZiff Davis Media eSeminars invite: Join this live event on April 26 at 2 p.m. ET and learn how to solve application performance problems and make the most of your organizations IT investment.

AppCritical is made up of an NIS (Network Intelligence System), typically installed on a server in the data center. It contains several analysis engines, a database and a user interface, and it manages probes or sequencers that are placed at strategic locations around the network.

A new monitor in the sequencers for AppCritical "keeps an eye on the microtests, knows about the paths, schedules continuous testing, and when it detects there is a problem, it accelerates the testing to verify there is a problem. Once it confirms the problem, it notifies the NIS to run a full test," Daniels said. The NIS evaluates the results of the full test and sends an e-mail or an SNMP notification or both to network operators.

With the tool, users can set thresholds for service quality definitions that allow users to ensure the network is providing the performance required by an application.

/zimages/3/28571.gifClick here to read about IBMs moves in the network management space.

Because AppCritical, which will cost about $100,000, can reduce the amount of time it takes to troubleshoot performance problems, beta tester Willis Marti looks at it as a "people multiplier," said the associate director for networking, computing and information services at Texas A&M University, in College Station.

"Its a continuous test that tells us when things go out of balance from what we expect or desire. Then AppCritical does its own shift into a detailed test and alerts us. By the time youve put a body on looking at a screen, youve got your first tests [done]. Then you can decide what to do after that point. So I can do more things with the same number of people because it points me to the right place," Marti said.

/zimages/3/28571.gifCheck out eWEEK.coms for the latest news, views and analysis on servers, switches and networking protocols for the enterprise and small businesses.