Total Virtual vCPUs
4,096 Cores
Usage: 64.2% Allocated
Memory Allocation
16.4 TB
Usage: 58.1% Allocated
Ingress Network I/O
84.2 Gbps
Zero Packet Jitter
HPA Autoscaler State
Dynamic OK
Target CPU: 70% Scale
Production Node Pools
Graviton4 & AMD EPYC 9654 Instances
pool-us-east-core (16x c7g.8xlarge)
Kubernetes v1.31 • 412 Pods • Kernel 6.8.0-aws
Ready • 48% CPU
pool-eu-db-primary (8x r7g.16xlarge)
NVMe High-IOPS • Distributed CockroachDB
Ready • 62% RAM
pool-gpu-inferencing (8x g5.12xlarge)
NVIDIA A10G vLLM TensorRT-LLM cluster
Ready • 89% GPU
Control Plane SLA: 99.999%
Prometheus Scrape Interval: 10s
ArgoCD Canary Rollout
Service: api-gateway-v2.14.0
Canary Traffic Weight
25% Progressive
Analysis Metrics:
• HTTP 5xx Error Rate: 0.00% (Pass)
• Latency p95: 14.2ms < 50ms (Pass)