An average poll lag time of a few seconds or more indicates your system may not be performing optimally. If WhatsUp Gold device polling seems to be experiencing performance lag, use the Poller Health dashboard report to assess and confirm poller performance. The WhatsUp Gold CPU and memory utilization reports can also be used to indicate performance issues. There are a number of ways to improve poller performance by reducing the workload of the WUG machine:
The first option is adding one or more additional pollers to your WhatsUp Gold system depending on the size of your network. When additional pollers are installed, load balancing should be disabled on the local poller using the procedure described below. This transfers the majority of the polling workload to the additional pollers, reserving the local poller for polling activity on the WhatsUp Gold Server. However, if your network is distributed across a large geographic area, you may benefit from assigning a poller to a specific subnet or device. In this case, load balancing should also be disabled on the specific poller to limit its activity to the assigned device(s).
The second option is removing the local poller from the load balancing queue reduces the workload of the WhatsUp Gold server and allows it to perform other tasks for which it is responsible.
To disable load balancing on the local poller:
The third option is to relocate your SQL instance to a machine separate from your WUG server. For more information, see the WhatsUp Gold Database Migration and Management Guide.
If you are still experiencing polling performance issues, consider the following network environment modifications: