Customers may observe frequent and unwarranted alarms from their external monitoring tool, reporting high memory and intermittent CPU utilization. These alerts are generated despite the NSX nodes operating within normal parameters.
VMware NSX
The issue arises due to the following reasons:
The alarm thresholds configured on the customer’s third-party monitoring tool are set too low compared to NSX’s native thresholds.
Even after upgrading the NSX version to 4.2.2, memory utilization for medium form factor nodes is expected to fluctuate between 85–88%, which is normal and within design specifications.
Current system usage shows memory levels below 88%, which is below NSX’s expected range. However, the external tool interprets these values as abnormal due to its overly sensitive threshold settings.
To resolve the false alarm issue, the following actions are recommended:
Adjust Thresholds on the External Monitoring Tool
Align the monitoring tool’s thresholds with NSX’s native levels:
Map the tool’s Warning threshold to NSX’s High threshold.
Map the tool’s High and Critical thresholds to NSX’s Very High threshold.
This adjustment ensures that alerts are triggered only when resource usage truly exceeds NSX’s expected operating levels.
Comparison Table
| Metric | External Tool Thresholds | NSX Recommended Thresholds | Recommended Mapping |
|---|---|---|---|
| CPU Usage | Warning: 75% High: 90% Critical: 92% |
High: 75% Very High: 90% |
Warning → High High & Critical → Very High |
| Memory Usage | Warning: 88% High: 90% Critical: 92% |
High: 88% Very High: 92% |
Warning → High High & Critical → Very High |
Alternative Option – Increase Form Factor
If the customer prefers to retain the existing low thresholds in their third-party monitoring tool, consider increasing the NSX Manager form factor to Large.
This provides additional CPU and memory resources, reducing the likelihood of hitting the current threshold values.