Meru Networks has released a new wireless LAN management application that for the first time runs on a dedicated appliance to collect only periodic summaries of access point statistics and look for problems.
Version 2.0 of E(z)RF Network Management digs up the equivalent of log information - packet-level data that shows the interactions "events") of the Wi-Fi client with access points and back-end servers. A Meru-built inference engine correlates the events, searching for patterns that show immediate or developing problems.
Company executives say the application can pinpoint specific problems quickly, and identify the root causes precisely. That in turn saves IT troubleshooting time, and minimizes or prevents users being offline and unproductive.
Problems, trends, WLAN states can be displayed in a package of customisable graphical dashboards that give network managers a quick visual summary of what's going on (an optional software module creates a visual representation by mapping this data to a floor plan showing WLAN access points and clients). With a few clicks, customers can dig down into details and histories of a given client or access point. Because this event data is stored in a PostgreSQL RDBMS on the appliance, managers can replay events to watch what happens when a laptop, for example, repeatedly requests an IP address.
The previous version of E(z)RF software, like most other WLAN management packages, periodically polled access points, requesting summaries of statistical data on throughput, signal strength and the like. Many enterprise sites have relied on a companion tool, a wireless sniffer that captures packet level data. But it's been difficult to bring together and correlate the information from the statistical tools and the sniffers, said Kamal Anand, senior vice president of marketing for Meru.
"There's been no concept of a history of [WLAN] events to see what was happening in the network at various time," he says.
Wireless events are a series of back and forth interactions that reveal how and why some feature or characteristic of a Wi-Fi client or access points changes. "You want to capture the meaningful bits of information, for example, those that cause a client's IP address to change," said Joe Epstein, Meru's senior director of technology. "We can capture these and see the actual error codes." It's this subset of data that the new E(z)RF version captures and uploads to the appliance-based inference engine for processing, analysis and correlation.
The result is a kind of automatic self-questioning management application that once it comes to a conclusion alerts the administrators about potential problems and their likely cause. In the past, a user's complaint of WLAN disconnects could require a largely manual, time-consuming process of eliminating possible causes, possibly even sending someone to check the access point. "With 2.0, if I see a client trying to associate [with an access point] many time, but they're not getting an IP address, the software triggers an alarm to check the DHCP server," Epstein said.
Meru has created a browser-based client for the 2.0 release, for the Apple iPhone and iPod Touch, dubbed E(z)RF OnTheGo. Users can connect wirelessly to the Meru management appliance over 3G or Wi-Fi links and use the mobile Safari Web browser to see the same set of dashboard displays as they would from a desktop or laptop PC. E(z)RF in turn can send alerts to the mobile user about high-priority events.
The E(z)RF Network Management rack-mounted appliance is $6,995. There are two software packages with pricing based on the number of access points. The base application, at $4,995 for 50 access points, provides all the core functionality - collecting, storing, correlating and reporting on WLAN event data. A second software module, also $4,995 for 50 access points, will map all this data to floor plans showing the locations of all Wi-Fi clients and access points.