Their process of setting details, thresholds and filters inside of System Center is a bit tedious. Although the tedium is somewhat assuaged by evolving administrator-defined groups of objects whose characteristics can be dealt en masse in a single action, such as aggregating, then setting bandwidth thresholds for VM Web server instances. Once this setup process is finished, using a console with two monitors can provide greater amounts of information to be displayed handily, and also permitted the busy System Center display to have a more comprehensive representation for the amount of information that's provided by nWorks MP.

A 'server state' view of System Center gives the best rapid visual indicator of overall systems health. As a tabular view of important information (including overall status, CPU threshold, network I/O, and other default or definable data), the system told us if all was well and healthy at a glance. We could then point to any VMware host, and generate a topological view of VM instances inside detailing their health characteristics.

NWorks gets information from the VMware Virtual Interface API and through VMware's VirtualCenter. We could find little difference in speed in the VM 'guest' instances between using VirtualCenter and Virtual Interface except when VM host utilization was at peak.

Stunning network and VM topological views can be easily configured, so that layouts can be viewed and understood. We created numerous custom templates with ease. Performance views are also easily made and enhance the value of twin monitors/displays to show comparative data graphically. The effect is very much like when Macs and Windows began to be used, and people would send memos with 20 fonts on them. Once under control, we found that sparse views of system health tabular data were the ones we referenced most.

The nWorks MP package, coupled to Microsoft's System Center, takes a long time to setup and configure, but pays off in monitoring control. Deciding thresholds and actions takes longer and requires a bit of practice. It's not a proactive package, and doesn't project thresholds like PlateSpin PowerRecon, and it has little knowledge of storage-area networks topologies, like Onaro SANscreen.

What we didn't like about nWorks MP mostly had to do with System Center. As an example, options to take actions that are nonsensical aren't disabled. One can take a silly action, like shutting down something that can't be shut down, only to have a nihilistic run at the attempt of shutting it down, generating a nice and inarticulate error report. PlateSpin and Onaro studiously avoid this. The nWorks MP also works with HP's OpenView, according to nWorks.

Overall, though, if Microsoft System Center is a management tool of choice, then nWorks MP for VMware makes a very good VM monitor and manager.