PULSE · Infrastructure health · API · DB · latency · uptime · Gemini anomaly watch
Overall Uptime
99.94%
API Latency
142ms
Active Connections
1,842
Incidents Today
1
API Gateway
42ms
p95 latency
99.99% uptime
DB Primary
8ms
query time
99.97% uptime
Kafka Broker
182ms
consumer lag
98.1% uptime
Redis Cache
94.2%
hit rate
99.95% uptime
Object Storage
99ms
avg upload
99.99% uptime
ML Inference
890ms
avg infer
96.8% uptime
Kafka consumer lag trending up
highMessage backlog will exceed 50k in ~18 min at current rate
Scale Kafka consumer group +2 instances
ML inference P99 > 800ms
highModel cold-start cascade likely if traffic +15% in next 30min
Pre-warm ML inference pool — 4 warm slots needed
DB connection pool at 78%
mediumPool exhaustion risk at peak traffic (+4h window)
Increase max pool size from 100 → 150 before peak
Object storage GET latency uptick
lowCDN edge cache stale — warming needed for Diwali campaign assets
Trigger CDN pre-warm for /assets/campaign/ prefix
Messages/sec
4,820
API calls/sec
892
DB queries/sec
2,140
Cache hit rate
94.2%