The alert is triggered when a physical host is in a powered-off state.
Host { .labels.target_hostname } is powered-off.
The host where the SCVM { .labels.target_hostname } resides is powered off.
The affected host cannot provide services. If HA is enabled for virtual machines running on the host, HA migration will be triggered to move the virtual machines to other hosts that meet their resource requirements and placement group policies. Data associated with the node will also be recovered on other nodes if the cluster has sufficient space and nodes.
The host is not powered on. Possible causes include:
Planned maintenance operations (scheduled shutdown and maintenance).
Unplanned shutdown due to an unexpected failure, for example:
Power outage: UPS failure or power load exceeding UPS capacity.
Hardware failures such as CPU, memory, or motherboard issues.
Planned shutdown
Typically, no additional action is required. Power the host back on after the maintenance is complete. The alert will be automatically cleared approximately 1 minute after the host starts normally.
Unplanned shutdown
Log in to the console via serial port or IPMI to check for power failure events or hardware alerts:
If a specific hardware failure is identified, contact the hardware vendor.
If no anomalies are found, check the cluster's power-on/off logs to determine whether a manual shutdown occurred.
If no hardware errors are found, try powering on the host:
If the system boots successfully and services start normally, the alert will clear. It is also recommended to check the host's kernel logs (/var/log/message) to confirm whether any hardware anomaly records exist.
If the system fails to boot normally, this typically indicates missing hardware (such as a system failure or unexpected hardware removal). Contact the hardware vendor for further investigation.
