Region
alicloud-ap-southeast-1
CPU Target
55% (+/-10% tolerance)
HPA Metrics
cpu (Resource) + ext_cpu_scaled_ratio (External)
Scaling Events
14
7 up / 7 down
Thrashing Rate
42.9%
3 of 7 events
Max P99 Latency
45.2s
avg 2.3s
Avg CPU Utilization
1.9%
max 67.2% / target 55%
Scoring Overview
Workload Timeline
Copr P99 Latency
TiDB P99 Latency
CPU Utilization
Replicas
Scale Out
Scale In
CPU Util Distribution vs HPA Zones
Detailed Scaling Events
| # | Time | Dir | Change | CPU Sparkline (+/-2min) | CPU@Event | Deviation | Threshold | Lag |
|---|
Total: 14 events
Scale-Up: 7
Scale-Down: 7
Scoring Breakdown
v0 Heuristic
72/100
Weighted Total: 74.75
5 categories evaluated
P99 Latency Distribution
Average
2.3s
Maximum
45.2s
Minimum
0.01s
Data Points
5,760
Normal
Acceptable
Degraded
Poor
Critical
HPA Configuration
CPU Target55%
Tolerance10%
Scale-Up Zone> 60.5%
Do-Nothing Zone49.5% - 60.5%
Scale-Down Zone< 49.5%
Min / Max Replicas2 / 30
Scale-Up PolicyMax
Scale-Down PolicyMax
Stabilization Window180s
CPU Request4 cores
Observed Replica Range2-5
Replica Distribution
Evaluation Criteria
v0 Heuristic Scoring: These criteria use rule-based heuristics. Scores are approximate and will be refined with ML-based evaluation in future versions.
Recommendations
Data Sources
Dataplane Prometheus
019bda6f-e0a4-7e29-bb26-5c4e0e874ec3
Application Prometheus
019bda70-1a2b-7c3d-aa45-6d5f1f985fd4
Generated: Feb 14, 2026 | Revision: v2.1 | Tool: scaling-evaluator | Methodology: autoscaler-dashboard v2
Namespace: tidb2013516099403907072