< Back | Next>
One of the VMs in your vSphere cluster is performing poorly. You suspect the VM either is not configured with sufficient resources or is contending with other VMs for available resources. Determine whether the issue is CPU or RAM related. Determine whether it is due to a lack of provisioned resources or due to contention.
- Use real-time performance charts (in the vSphere Client or vSphere Web Client) or resxtop to determine whether the issue is one of the following:
- The VM is contending too much with other VMs for CPU usage.
- The VM is contending too much with other VMs for RAM usage.
- The VM needs additional configured virtual CPUs.
- The VM needs additional configured virtual RAM.
- If the VM is contending too much with other VMs, shut down a noncritical VM.
- If the VM needs additional virtual CPUs, shut down the VM and add one virtual CPU. Leave the VM shut down.
- If the VM needs additional virtual memory, shut down the VM and increase its virtual memory by 1 GB. Leave the VM shut down.
- VM name: vm-10
- Non-critical VM name: vm-11
VDCA510 Objective: 6.2 Troubleshoot CPU and Memory Performance
VDCA550 Objective: 4.2 Optimize Virtual Machine Resources
- Chapter 4 - Troubleshoot CPU and Memory Performance
- Chapter 10 – Scenario 10-17
Click Here For Solution
- Use resxtop or performance graphs (real-time) to display CPU statistics for the VM.
- Examine the current CPU ready time for the troubled VM. If the value is less than 5% in resxtop and less than 1000 ms in the performance graphs, the problem is not CPU contention.
- Examine the current memory ballooning, compression, and swapping values for the VM. If the values are zero for each item, the problem is not RAM contention.
- Examine the CPU usage. If it is consistently near 80% or more or if it frequently spikes to 90% or more, it might benefit from additional vCPU.
- Examine the memory usage of the VM. If it is consistently near 80% or more or if it frequently spikes to 90% or more, it might benefit from additional vRAM.