Argent Global Alert

An NT system watchdog

Do you scroll through Windows NT's event logs looking for nasty little presents from your operating system (OS) every day? Would you rather have NT automatically monitor its logs and alert you if a serious problem occurs? If you do, Argent Global Alert (AGA) from Argent Software is the product you need.

AGA is a complete system that monitors performance, event logs, and service. AGA can scan through your NT event logs and monitor other critical system features. When the software detects a problem, it automatically alerts you through email, alphanumeric or numeric paging, or the NT messenger service. Screen 1 shows the alert methods you can select.

AGA can use NT's management feature to monitor your entire NT network from one server. AGA doesn't require special OS hooks, device drivers, or other NT subsystem changes.

You must create and maintain three sets of values to work with AGA: rules, node lists, and alerts. A rule defines a condition you want to monitor, such as your system's disk space (performance rule) or a message in your event log (event rule). You can also monitor the status of your system's services (service rule) and your system's host availability (system-down rule). A node list specifies which computers must follow the rules. Finally, AGA generates an alert when your system trips one of the node's rules (e.g., if you exceed the preset disk-space limit).

To use the software, you first create a set of rules, node lists, and alerts. Next, you associate a rule, a node list, and an alert to create a condition for the software to monitor. When the machines on the node list meet the conditions associated with the rule, AGA processes the alert to notify the systems administrator.

Installing the software requires minimal effort. My test copy arrived on a 3.5" disk. When I ran the setup program, it copied several files to my system. The software then installed a service in my configuration to automatically start AGA whenever my NT system booted.

To test the software, I created a performance rule to monitor disk space. As part of the rule, I provided the system parameter (logical disk 0) that I wanted to monitor and the value the parameter had to fall below (5 percent) for AGA to generate a message.

As an alternative test, I decided to use the system-down detection capability (i.e., Dwn Rule). From the Start menu, I selected the Dwn Rule icon within the AGA Manager. The software presented a screen on which I needed to specify how I wanted the software to test for system availability. Because all my nodes are TCP/IP (not Windows machines only), I selected the ping option.

For both tests, I had to create a node list and an alert. I used two node lists: one for my NT Server machine (for the disk-space monitoring) and one with NT, Windows 95, and UNIX machine names. After I associated the rules with the alerts and a node list, AGA began monitoring my system and performed the necessary correction when my system triggered an alert.

AGA is an excellent product, and it's indispensable in mission-critical NT environments. However, AGA is too expensive to be practical for most users. The base AGA console includes 10 event log scanning engines, service monitoring engines, performance statistics nodes, and system alive engines. If Argent provided the software in a smaller configuration (e.g., let administrators monitor one small server), AGA might be more affordable. But unless you have a large NT environment, leave the product where it is: in the vendor's inventory.