List VM's Metrics

GET/v1/compute/vms/{vm_id}/metrics

Read a VM’s resource metrics over an interval. Every series covers the same periods, so they line up index for index, and a period the VM reported no observation for carries a null value.

Path ParametersExpand Collapse

vm_id: string

Query ParametersExpand Collapse

aggregation: optional "mean" or "max" or "min"

How the samples inside one period are folded into a single value.

One of the following:

"mean"

"max"

"min"

end_time: optional string

End of the interval, exclusive, as an RFC 3339 timestamp. Defaults to now.

formatdate-time

metric: optional array of "compute.nirvanalabs.io/vm/cpu/used_cores" or "compute.nirvanalabs.io/vm/cpu/utilization" or "compute.nirvanalabs.io/vm/memory/used_bytes" or 7 more

Metric to return. Repeat the parameter for several; every metric is returned when it is left out.

One of the following:

"compute.nirvanalabs.io/vm/cpu/used_cores"

"compute.nirvanalabs.io/vm/cpu/utilization"

"compute.nirvanalabs.io/vm/memory/used_bytes"

"compute.nirvanalabs.io/vm/memory/utilization"

"compute.nirvanalabs.io/vm/disk/read_bytes"

"compute.nirvanalabs.io/vm/disk/write_bytes"

"compute.nirvanalabs.io/vm/disk/read_ops"

"compute.nirvanalabs.io/vm/disk/write_ops"

"compute.nirvanalabs.io/vm/network/rx_bytes"

"compute.nirvanalabs.io/vm/network/tx_bytes"

period: optional "5m" or "15m" or "1h" or 2 more

Width of one period, and so the spacing between points. An interval holding more than 1440 periods is rejected; the error names a period that fits.

One of the following:

"5m"

"15m"

"1h"

"6h"

"24h"

start_time: optional string

Start of the interval, inclusive, as an RFC 3339 timestamp. Defaults to an hour before end_time. A start older than the 30 days of history kept is served from where that history begins.

formatdate-time

ReturnsExpand Collapse

VMMetrics object { aggregation, end_time, metrics, 3 more }

A VM’s metrics over an interval: one series per metric, on a shared grid of periods.

aggregation: "mean" or "max" or "min"

How the samples inside one period were folded into a single value.

One of the following:

"mean"

"max"

"min"

end_time: string

End of the interval served, exclusive.

formatdate-time

metrics: array of VMMetricSeries { metric, points, unit }

One series per requested metric, in the order they were requested.

metric: string

Fully-qualified name of the metric.

points: array of VMMetricPoint { timestamp, value }

Values over the interval, oldest first. Every series in a response covers the same periods, so they line up index for index.

timestamp: string

End of the period the value covers, so a point timestamped 00:05 over five-minute periods describes 00:00 through 00:05.

formatdate-time

value: number

Value over the period, in the series’ unit. Null means the VM reported no observation for this period, which is what a stopped VM looks like.

unit: VMMetricUnit

Unit the values are expressed in.

One of the following:

"ratio"

"bytes"

"cores"

"bytes_per_second"

"operations_per_second"

period: "5m" or "15m" or "1h" or 2 more

Width of one period, and so the spacing between consecutive points.

One of the following:

"5m"

"15m"

"1h"

"6h"

"24h"

start_time: string

Start of the interval served, inclusive. It can be later than the requested start when the request asked for more than the available history.

formatdate-time

vm_id: string

ID of the VM the series belong to.

List VM's Metrics

curl https://api.nirvanalabs.io/v1/compute/vms/$VM_ID/metrics \
    -H "Authorization: Bearer $NIRVANA_LABS_API_KEY"

{
  "aggregation": "mean",
  "end_time": "2026-01-01T01:00:00Z",
  "metrics": [
    {
      "metric": "compute.nirvanalabs.io/vm/cpu/utilization",
      "points": [
        {
          "timestamp": "2026-01-01T00:05:00Z",
          "value": 0.42
        }
      ],
      "unit": "ratio"
    }
  ],
  "period": "5m",
  "start_time": "2026-01-01T00:00:00Z",
  "vm_id": "123e4567-e89b-12d3-a456-426614174000"
}

Returns Examples

{
  "aggregation": "mean",
  "end_time": "2026-01-01T01:00:00Z",
  "metrics": [
    {
      "metric": "compute.nirvanalabs.io/vm/cpu/utilization",
      "points": [
        {
          "timestamp": "2026-01-01T00:05:00Z",
          "value": 0.42
        }
      ],
      "unit": "ratio"
    }
  ],
  "period": "5m",
  "start_time": "2026-01-01T00:00:00Z",
  "vm_id": "123e4567-e89b-12d3-a456-426614174000"
}