Virtual Machine Management (VMM)

Virtual Machine Management (VMM) metrics track CPU, memory, disk, and network usage for virtual machines (VMs) in your Google Distributed Cloud (GDC) air-gapped projects. Your default permissions let you view metrics for the VMs in your projects. Platform administrator (PA) permissions also grant access to metrics for user workload VMs and Kubernetes cluster node VMs across all projects. GDC doesn't support guest environment metrics on Windows VMs.

For information about the VM-level metrics dashboard and how it visualizes these metrics, see Monitor VM metrics.

Metrics list

Use the namespace label to filter the VMs in different namespaces.

Name and Description Type Scrape interval Labels
compute_instance_cpu_usage
The amount of time each vCPU spends in different states. Counter 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
  • state: The state of the vCPU, such as running, stopped, or paused.
  • id: The numerical index of the vCPU, which ranges from 0 to the maximum vCPU count - 1.
compute_instance_memory_available_bytes
The amount of memory available to the VM. Gauge 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
compute_instance_memory_usable_bytes
The amount of memory that you can give to the VM from the available memory without having to use the swap or ballooning. Gauge 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
compute_instance_memory_total_bytes
The amount of memory that the VM actually uses. Gauge 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
compute_instance_disk_bytes_allocated
Count of bytes totally allocated to the disk from the underlying storage system. Gauge 1 min
  • name: The name of the VM disk.
  • namespace: The project to which the VM disk belongs.
  • pvc: The underlying PersistentVolumeClaim name.
compute_instance_disk_bytes_used
Count of bytes that the disk uses in the underlying storage system. This metric reflects on total disk storage being utilized. However, the metric lacks visibility into the actual file system usage, which is only known to the guest operating system. Gauge 1 min
  • name: The name of the VM disk.
  • namespace: The project to which the VM disk belongs.
  • pvc: The underlying PersistentVolumeClaim name.
compute_instance_disk_read_bytes_count
Count of bytes read from disk. Counter 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
  • drive: The name of the VM disk drive.
compute_instance_disk_write_bytes_count
Count of bytes written to disk. Counter 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
  • drive: The name of the VM disk drive.
compute_instance_disk_read_ops_count
Count of disk I/O read operations. Counter 10 s
  • name: The name of the VM.
  • namespace: The project associated the VM.
  • drive: The name of the VM disk drive.
compute_instance_disk_write_ops_count
Count of disk I/O write operations. Counter 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
  • drive: The name of the VM disk drive.
compute_instance_disk_read_time_ms
The cumulative time spent on the read operations that are in progress. That is, the actual time in queue and when disks are busy. Counter 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
  • drive: The name of the VM disk drive.
compute_instance_disk_write_time_ms
The cumulative time spent on the write operations that are in progress. That is, the actual time in queue and when disks are busy. Counter 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
  • drive: The name of the VM disk drive.
compute_instance_disk_flush_requests_total
Storage flush requests. Counter 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
compute_instance_disk_flush_times_ms
Total time in microseconds spent on cache flushing. Counter 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
compute_instance_network_received_bytes_count
Count of bytes received from the network. Counter 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
  • interface: The network interface name, such as eth0.
  • network: The name of the network. For example, pod-network.
compute_instance_network_received_packet_count
Count of the packets received from the network. Counter 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
  • interface: The network interface name, such as eth0.
  • network: The name of the network. For example, pod-network.
compute_instance_network_sent_bytes_count
Count of bytes sent to the network. Counter 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
  • interface: The network interface name, such as eth0.
  • network: The name of the network. For example, pod-network.
compute_instance_network_sent_packet_count
Count of packets sent to the network. Counter 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
  • interface: The network interface name, such as eth0.
  • network: The name of the network. For example, pod-network.
compute_instance_network_receive_errors_count
Count of error packets received. Counter 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
  • interface: The network interface name, such as eth0.
  • network: The name of the network. For example, pod-network.
compute_instance_network_receive_packets_dropped_count
The number of RX packets dropped on vNIC interfaces. Counter 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
  • interface: The network interface name, such as eth0.
  • network: The name of the network. For example, pod-network.
compute_instance_network_transmit_errors_count
Count of network error packets transmitted. Counter 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
  • interface: The network interface name, such as eth0.
  • network: The name of the network. For example, pod-network.
compute_instance_network_transmit_packets_dropped_count
The number of TX packets dropped on vNIC interfaces. Counter 10 s
  • name: The name of the VM.
  • namespace: The project associated with the VM.
  • interface: The network interface name, such as eth0.
  • network: The name of the network. For example, pod-network.