cloud-monitoring-list-time-series-request

Provide the metric type, resource type, and time interval, then construct a filter with metric.type and resource.type clauses, choose perSeriesAligner and crossSeriesReducer based on the metric's kind and value type, and set alignmentPeriod from the lookback duration.

GAUGE numeric metrics default to ALIGN_MEAN with REDUCE_SUM, DELTA or CUMULATIVE counters use ALIGN_RATE with REDUCE_SUM, and DISTRIBUTION metrics use ALIGN_DELTA or percentile aligners. Utilization ratios must never use REDUCE_SUM across instances.

Use metric.labels.instance_name or metadata.system_labels.name for human-readable GCE instance names, since resource.labels.instance_id expects a numeric ID. Never use resource.metadata.name, which is invalid in Cloud Monitoring filter syntax.

Set alignmentPeriod based on lookback duration: 60s for up to 110 minutes, 300s up to 23 hours, 3600s up to 6 days, scaling to 172800s for up to 500 days. Omit it only when perSeriesAligner is ALIGN_NONE.

Do not use it for metric discovery or metric selection; use the cloud-monitoring-metric-selection skill first when the target metric type is unknown. It also requires a resolvable GCP project ID before generating requests.