ISPPlanet Network Management System Series - ipMonitor

Scheduled Jobs and Custom Reports

Given ipMonitor's event-driven infrastructure, scheduling jobs (right) is a natural extension. Jobs are simply task sequences, invoked daily, weekly, monthly, or every N days. Jobs can send scheduled reports, execute third-party software or batch files, and restart NT services on a regular basis. A "wait" task can be used to suspend job processing for a few seconds, until a service is running or a device becomes reachable.

NT Service Control

When ipMonitor runs under an account that has been granted access to NT services, the GUI menu bar includes "WinNT Control" options. List Network browses the Network Neighborhood; Select Server finds a computer by NetBIOS name or address. ipMonitor displays all NT services (right) configured on the target workstation and current run status. Services can be started or stopped by clicking on service icons.

These panels provide remote control over NT services, but also come in handy when defining Monitors or Alerts that operate on NT services. Why? Both require workstation (UNC) and service names are easily viewed through these panels. If you are unable to view a server or service through WinNT Control, you won't be able to successfully monitor or restart the service or reboot the server with ipMonitor. We used this to diagnose a failed SERVICE Monitor automatically created by Add Network. The Monitor was misconfigured with a fully-qualified hostname rather than UNC name. Using WinNT Control, we verified our ability to query the server by UNC name, then reconfigured the SERVICE Monitor to match.

Status Reporting
All configuration and control features are accessed through ipMonitor's "Config" GUI. But once ipMonitor is up and running, most of your interaction will involve ipMonitor's "Reports" GUI.

The page you'll probably watch all the time is Monitor Summary (left), a self-refreshed status display by Group. Failed dependency turns the Group and Group /Dependency column red, any failed member turns the Group/Member column red. Yellow shows potential failure—for example, a Monitor that polls three times before generating an Alert turns yellow after the first and second poll, red after the third. Grey indicates administratively-suspended Monitors.

Click through the Summary page to display Monitor status (right): just Monitors in a Group or everything known to ipMonitor, in tabular or block (detail or control room) formats. At this drill-down level, Monitor name, type, current status, cumulative up/down time, and next poll time are displayed. Here you can click on a Monitor's name to see its current configuration. Nice additions would be click-through to Alert activity (how many alerts, of which type, generated for this failure) or to view a report on the affected Monitor.

History Status displays a trio of log files. Monitor history lists the date and time of each status change. Notification (left) history shows the Profiles and Alerts checked when a Monitor fails. This log is useful to verify Profile behavior, but should also include Alert result. You'll need good eyesight—fonts are tiny. Thankfully, history logs can be filtered by timestamp, name, status, action, etc. A Security log indicates when other files are reset or archived.

To assess long-term availability, downtime, response time, and failure events (troubles), ipMonitor can generate statistical reports (right) for a selected Monitor or Group. Display a list of Monitors by device or Group to view a generic report, similar in content and structure to the custom reports initiated by Jobs and Alerts. A "Change Report" panel can be use to customize the report on the fly (e.g., change time period, report style, graphing options).

Final Words
ipMonitor is a simple way to continuously watch applications and devices, record on-going statistics, log transient failures, and be notified about persistent failures. Groups provide data reduction, enabling "at a glance" status and reducing alerts when a single failure has many side-effects. We believe ipMonitor is suitable for granular surveillance in small networks or selective surveillance in other networks. (At some point, the number of Monitors become unwieldy.)

ipMonitor's flexibility make it useful many environments: ISPs can monitor routers and firewalls, ASPs can monitor application services. But it's important to understand ipMonitor's boundaries: you'll need to write your own scripts for *NIX server recovery and there's no alert acknowledgement or progress tracking. Although ipMonitor includes SNMP polling and TRAP monitoring, it cannot be used to perform ad hoc SNMP GET queries or reconfigure managed devices using SNMP SET—it's a monitor, not a manager. One shouldn't expect ipMonitor to provide large scale, carrier-class network management. But, at $695, it packs a lot of punch for small-to-midsize network monitoring.

ipMonitor v6.02 $695 MediaHouse Software, Inc. Hull, Quebec, J8Y 1V4, Canada http://www.mediahouse.com