Bake-Off: 3 Network Management Solutions Put To The Test

AdRem NetCrunch 4.3
AdRem Software's NetCrunch 4.3 did something neither of its peers did: It automatically discovered the two VMware virtual servers that were on the Windows XP workstation. Test Center engineers had forgotten the presence of the virtual servers and were both surprised and impressed to see them detected by NetCrunch.

Overall, NetCrunch turned in solid performance and would satisfy the management needs of most small-to-midsize businesses, especially those running primarily Windows servers. NetCrunch delivers in the areas that are most critical to network management, such as device discovery, resource monitoring, events and alerts, trending and diagnostics. Engineers noted, however, that while the product works with non-Windows systems and SNMP devices, some tweaking was required to match the level of information that is automatically available for Windows machines. Fortunately, NetCrunch's template editor is easy to use.

Developing templates to manage non-Windows platforms could be a place for solution providers to add value and show expertise.

Engineers installed NetCrunch 4.3 Premium XE instead of NetCrunch Premium. The more scalable XE package is intended for complex monitoring on networks with several hundred nodes. Like its peers, it can map the network topology and collect bandwidth data. NetCrunch creates an "atlas" of all known networks and nodes. The atlas can be created automatically using predefined requirements or by specifying the types of devices to include. Engineers defined and ran a series of IP scans to populate the atlas database and viewed the dynamic "map" views.

Like its peers, NetCrunch uses sequential ICMP ping messages, SNMP 1, 2 and 3 queries, NetBIOS and WMI lookups to auto-detect devices. Along with the aforementioned VMware servers, NetCrunch also can discover devices registered in Novell eDirectory servers. Wizards help define more complicated scans and allow specific IP addresses to be added manually.

Engineers were able to look at graphical representations of bandwidth usage for each Linux and Windows machine on the network, as well as perform traceroutes. Engineers polled services, event logs and system hardware information on the Windows machines. NetCrunch uses ping to measure availability to monitor a handful of TCP services. It also can read a variety of performance data from SNMP and WMI queries, as well as listen for incoming SNMP traps and SYSLOG messages. The graphs are single-element charts but are adequate for most management needs. Reports can be generated from performance readings and graphs to show current and long-term trends.

NetCrunch includes a set of basic tools for troubleshooting, such as traceroute, MIB and WMI browsers, and DNS debuggers. It can pop up messages on the management console or send text or e-mail alerts under specified conditions. Like LANSurveyor, NetCrunch can shutdown or restart a machine remotely, but it takes its remote abilities one step further: It can manipulate Windows services as well, so problematic services can be restarted without rebooting the entire machine. NetCrunch also can be configured to log in to a remote system over SSH and upload and run a custom script.

Like Ipswitch's WhatsUp Gold, NetCrunch supports role-definable accounts on its Web-based management interface. So while the application and its data-collection engine runs only on Windows, users can log in from other consoles to manage the network.

Next: The Bottom Line