![]() |
NetIQ
AppManager
by Bob Walder
Windows-based networks have grown in size and complexity over the years, and many larger organisations have a number of distributed, mission-critical applications running on the Microsoft flagship OS. Managing NT can be a nightmare at the best of times, especially when servers are geographically dispersed. So how can a network administrator be expected to handle all those distributed servers layered with complex applications such as SQL Server, Oracle, IIS, ARCserve, MTS and so on? And what about the hardware on which they are running? Manufacturers such as Compaq and Dell include management agents and software to monitor their own hardware platforms, but trying to manage competing platforms from a single location is fraught with problems. Help could be at hand with NetIQ AppManager Suite, a comprehensive solution for managing and monitoring the performance and availability of distributed NT and Windows 2000 systems, applications and hardware. AppManager allows system administrators to view all of their Windows NT and Windows 2000 servers and workstations from a single, central console and to monitor computer resources and application configuration, check for potential system problems, initiate responsive actions, and gather performance data for real-time and historical reporting. AppManager can be a complex beast to buy, let alone install and configure, since it consists of a number of different product options that can be used together or separately, depending on your requirements. Each product option in the Suite is specially focussed to perform specific management tasks for a particular application or resource. For instance, in the Windows NT/2000 category there are agents to cover the base operating system, Citrix WinFrame, Active Directory, Cluster Server, Network Load Balancing, and Terminal Server, and out-of-the-box monitoring scripts provide the ability to identify CPU bottlenecks and terminate runaway processes, check if key services are down and auto start them, track disk space usage, reboot downed servers, determine if running low on DHCP leases, and so on. If it is messaging systems that interest you, AppManager covers Lotus Domino, Exchange Server and Microsoft Message Queue Server, performing such tasks as monitoring e-mail connectivity and response time, reporting on e-mail traffic flow, identifying top e-mail senders and receivers, monitor e-mail disk space usage, and restarting services or tasks that have failed. Other categories covered include Web servers (IIS, Proxy Server, Site Server), database servers (SQL Server, Microsoft Transaction Server, Oracle, SAP/R3), hardware (Compaq Insight Manager, Dell OpenManage, HP TopTools for Servers, IBM NetFinity Manager, Siemens ServerView), tools and utilities (CA ARCserve, Legato Networker, SNA Server, SMS, Veritas Backup Exec), and connectors (Cabletron Spectrum, CA Unicenter TNG, HP OpenView NNM, Tivoli Enterprise), each with its own particular set of monitoring capabilities. Basically, AppManager uses a number of pre-defined rules known as Knowledge Scripts in order to perform one or more management tasks, such as monitoring CPU usage, disk usage, verifying connectivity between messaging hosts, or detecting whether or not a critical service is up and running.�
Depending on the task, Knowledge Scripts can collect performance data (such as CPU usage), monitor systems for simple or complex events (such as whether a device or service is down), and respond with one or more actions (such as paging or e-mailing the administrator, or automatically re-starting a failed service). New knowledge scripts can be created by the administrator if required, and Script parameters can be quickly and easily changed to reflect new or changing requirements. The AppManager multi-tiered architecture is designed to be scalable, in order to provide central monitoring and management for an entire organisation. At the lowest level are the Agents, comprising Windows NT/2000 services that perform the actual monitoring of the host systems, communicating their results to the Management Server. The Management Server manages the event-driven communication from the Agents and stores the results in the Repository, a SQL Server database where all AppManager data resides for alerting and reporting purposes. Finally, there is the AppManager Console, the GUI interface where the administrator can configure the monitoring services and view alerts and reports (via the AppManager Report Manager). Web versions of the Console and Management Server are also available. The Console presents an initially daunting multi-paned view to the administrator which, after an initial network discovery process during which all active Agents are located, consists of a hierarchical Explorer-like display of managed objects, along with all the systems resources and applications located on each of those objects. Thus it is possible to expand the entry for an individual machine and examine critical resources such as CPU activity and disk usage, as well as application-specific data such as Exchange log file usage. It is also possible to select a number of alternative application-oriented views which displays all objects that are running Exchange Server, IIS, SQL Server, and so on, and individual resources can be arranged into logical groups for ease of administration. Jobs are created by dragging and dropping Scripts onto the managed objects in the AppManager Console, where they can be scheduled for once-only or regular runs. One parent job can spawn any number of child processes, thus making it easy to change job parameters for hundreds of machines at a time, and then apply them in a single operation. Network traffic is kept to a minimum by ensuring that only key exceptions and final results are reported back to the Management Server, rather then maintaining constant communication. Clicking on any managed resource will show the jobs currently running or scheduled to run in the future on that object, and results can be displayed in an easy-to-understand graph or report format. Verdict� AppManager is necessarily complex, since it is designed to monitor and manage complex network environments. However, once you have it configured and running it makes the day-to-day management of an NT/Windows 2000-based network as simple as it is ever likely to get. Well worth a look. Contact:
Peapod Group
|
![]() |
Send mail to webmaster
with questions or�
|