3.5.1 Analyzing the cause

The following describes how to analyze the cause.

Organization of this subsection

(1) Analyzing whether the bottleneck is caused by the usage of the business system

Procedure

In the Analyze Bottleneck window, click Check Noisy Neighbor.

The window for analyzing the cause of a bottleneck appears.

Bottleneck Candidate displays the graph for the CPU that was determined to be the bottleneck. Resources using the Bottleneck displays the graphs for VMs.
Select CPU Ready from Metric of Bottleneck Candidate, and then select Virtual Machine CPU Use from Metric of Resources using the Bottleneck. Then, compare the CPU Ready value with the CPU usage to identify the VM that is placing a load on the CPU of ESX.

If the waiting time of the virtual machine was increased at almost the same time as the increase in the CPU usage, there is a problem with the usage of the business system.
Look for a graph that has the same shape as the CPU.

Because there are no graphs that have the same trend, the usage of the business system did not cause the bottleneck.

Tip

If you cannot determine the cause of the bottleneck from the displayed graphs, check the graphs of different metrics. To view a graph of a different metric, select the metric from Metric or add a new graph by selecting the Add Graph menu of Resources using the Bottleneck, and then select the metric from Metric.

Next steps

Analyze whether the bottleneck was caused by changes in the configuration of the business system.

Procedure

In the Analyze Bottleneck window, click Check Related Changes.

A bar graph indicating the number of events related to changes in the business system configuration, and a line graph for the CPU that was determined to be the bottleneck are displayed.
Click the bar graph for the time period before or after the time when the line in the graph for the CPU rises.

The table displays previous changes that occurred in the time period shown in the graph as events.
Check the details of the events related to changes in the business system configuration in the table.

Check whether an event related to changes in the business system configuration affected the bottleneck.
By checking the details for the event in which a VM was added, you can determine that the new VM caused the bottleneck.
Write down the name of the VM that was added, and then close the Analyze Bottleneck window.
In the search box in the Workspace window, enter the VM name that you checked in step 4.
Select Servers as the type, and then click Search.

The VM is displayed.
Select the VM.
Click Show E2E View.
Click the icon for the VM. From the displayed menu, select Show Detail to check the performance.

Because the performance of the VM affects ESX, ESX can no longer manage the VM. This means that there is a problem in the business system configuration.

Next steps

Because you were able to recognize that the addition of a VM caused the bottleneck, give directions to remove the cause.