These steps help you identify and resolve the issues caused by burstable nodes.

1. Review your resource requests and limits
   Make sure your pods define realistic CPU and memory requests and limits. Burstable nodes perform inconsistently when pods don’t request enough guaranteed resources. You can use Kubernetes resource quotas and limit ranges as a reference when setting these values.
   
   

2. Exclude B‑series nodes from your virtual node groups
   If your workloads require consistent performance, remove B‑series nodes from your Ocean virtual node groups so your pods run on non‑burstable instance types.
   
   

3. Use rightsizing recommendations
   Review and apply Ocean rightsizing recommendations to ensure your pods request the correct amount of CPU and memory for stable operation.