How do I interpret these results?
Capacity Planner simulation results can help you determine whether over-utilized or under-utilized hardware is affecting your capacity model and whether there are any high-latency transactions in your capacity model.
Note
If you think you might have resources that are under-utilized or over-utilized or if you think transactions have unacceptably high latency, see Over-Utilized Resources and High-Latency Transactions.
After you run the simulation, by clicking Run Simulation on either the Model Wizard or the Model Editor toolbar, the Capacity Planner Results Summary page appears. This page displays the following sections:
- Bottleneck analysis This bar graph depicts hardware devices or network connections in the capacity model that are the most heavily used, in terms of percentage of total usage.
- Highest storage space utilization This bar graph displays the highest space utilization, per storage device.
- Longest transactions (sec) This list shows transactions that take the longest amount of time to complete.
- Highest CPU utilization This list shows processors or processor groups have the highest processor utilization, in percentage of total usage.
- Highest storage I/O utilization This list shows disk volumes that have the highest disk utilization, in percentage of total usage.
- Highest connectivity utilization This list shows network connections that have the highest utilization, in percentage of total usage.
Note
The last four sections in the preceding list display threshold indicators, which are represented by icons on the Results Summary page that indicate how hardware and network resources are utilized in a capacity model. These icons vary in color depending on whether hardware or network resource values do not exceed a threshold, are within 5 percent of a threshold, or exceed a threshold that you have set for hardware devices on the Threshold settings page.
For transactions that show high latency or take a long time to complete, as indicated by the threshold indicators in those sections of the Results Summary page, you can try to reduce the latency by doing the following:
- Increase resources available to the computer where the transaction runs.
- Move the components that generate the transactions to a less heavily loaded computer.
- Experiment with alternate topologies and routing strategies.
For a hardware or network resource that exceeds the threshold for a device, as indicated by the threshold indicators in those sections of the Results Summary page, you can try to reduce the utilization by doing the following:
- Reduce the usage either by changing the usage parameters (for example, reducing the number of users) or by redirecting some traffic to other devices.
- Reduce the load on the device by changing the computer configuration to add more devices of the same type.
- Change the device configuration to handle additional load.
- Reallocate a portion of the load to another server by moving the server to a different location or by adding another server to redistribute the load.
- Increase the capacity of a WAN by adding additional WAN links.