CPU Throttling Under Query Load

critical

Resource ContentionUpdated Feb 23, 2026

Container CPU throttling increases when vector search query volume spikes, indicating the cluster is CPU-constrained and unable to meet query demand efficiently.

Sources

Monitor Clusters - Qdrantqdrant.tech

Monitoring & Telemetry - Qdrantqdrant.tech

Technologies:

QdrantSymptoms of this issue are visible in Qdrant metrics and logs

KubernetesThe root cause of this issue originates in Kubernetes

How to detect:

Monitor container_cpu_cfs_throttled_periods_total counter for increases correlated with qdrant_collection_number_of_grpc_requests or qdrant_collection_number_of_rest_requests spikes. CPU throttling indicates demand exceeds allocated CPU.

Recommended action:

Scale cluster vertically by increasing CPU limits in kube_pod_container_resource_limits, or horizontally by adding replicas. Review query patterns to optimize top_k values and filter complexity.