Troubleshooting resizing recommendations for overallocated VMs
The following table explains the accuracy levels that are associated with the resizing recommendations for overallocated VMs. You can use the information in the Probable reason and Probable resolution columns to understand the cause of the levels and the actions that you can take to improve them.
Accuracy of overallocated VM recommendations
The Overallocated VM recommendation window displays an Accuracy icon that indicates the accuracy or reliability of the resizing suggestion given in the recommended actions. For more information about how the recommendation is generated, see Overallocated VM recommendation.
The Accuracy icon can have one of the following values:
Icon - Accuracy level
- Medium
- High
- Very High
Recommendations and probable resolutions
| Icon/ Accuracy level | UI message | Probable reason | Probable resolution |
|---|---|---|---|
| Although balanced behavior was selected, the recommendation is based on aggressive behavior because detailed statistics for hourly data samples are not available for the server. | The CPU_UTILMHZ_HM metric is missing. Utilization metrics of the source server are collected at hourly granularity. About CPU_UTILMHZ_HM metric The CPU_UTILMHZ_HM metric is the CPU utilization value (expressed in MHZ) that is calculated by using granular samples that the Continuous Optimization Agent collects. For balanced behavior, the 90th percentile value of CPU utilization is considered. For conservative behavior, the 95th percentile value of CPU utilization is considered. |
|
| Although conservative behavior was selected, the recommendation is based on aggressive behavior because detailed statistics for hourly data samples are not available for the server. | The CPU_UTILMHZ_HM metric is missing. Utilization metrics of the source server are collected at hourly granularity. About CPU_UTILMHZ_HM metric The CPU_UTILMHZ_HM metric is the CPU utilization value (expressed in MHZ) that is calculated by using granular samples that the Continuous Optimization Agent collects. For balanced behavior, the 90th percentile value of CPU utilization is considered. For conservative behavior, the 95th percentile value of CPU utilization is considered. |
|
| Although balanced behavior was selected, the recommendation is based on 15 minutes detail samples. Detailed statistics for hourly data samples are not available for the server. | The CPU_UTILMHZ_HM metric is missing. Utilization metrics of the source server are collected at 15-minutes granularity.
About CPU_UTILMHZ_HM metric The CPU_UTILMHZ_HM metric is the CPU utilization value (expressed in MHZ) that is calculated by using granular samples that the Continuous Optimization Agent collects. For balanced behavior, the 90th percentile value of CPU utilization is considered. For conservative behavior, the 95th percentile value of CPU utilization is considered. |
|
| Although conservative behavior was selected, the recommendation is based on 15 minutes detail samples. Detailed statistics for hourly data samples are not available for the server. | The CPU_UTILMHZ_HM metric is missing. Utilization metrics of the source server are collected at 15-minutes granularity.
About CPU_UTILMHZ_HM metric The CPU_UTILMHZ_HM metric is the CPU utilization value (expressed in MHZ) that is calculated by using granular samples that the Continuous Optimization Agent collects. For balanced behavior, the 90th percentile value of CPU utilization is considered. For conservative behavior, the 95th percentile value of CPU utilization is considered. |
|
| The recommendation is based on aggressive behavior, as requested. | The metrics that are required for the selected behavior are available and the recommendation is generated as expected. | No action is required. |
| The recommendation is based on balanced behavior, as requested. | The metrics that are required for the selected behavior are available and the recommendation is generated as expected. | No action is required. |
The recommendation is based on conservative behavior, as requested. | The metrics that are required for the selected behavior are available and the recommendation is generated as expected. |
Information: The accuracy level is High as compared to the earlier scenarios because the 15-minute data samples are available as compared to hourly granularity.
Comments
Log in or register to comment.