Advanced Performance Analysis & Capacity Planning
Accounts for GC, OS scheduling, context switches
How much traffic spikes above average
Additional capacity from hyperthreading
L = λ × W
L = Concurrent requests
λ = Arrival rate (TPS)
W = Service time (latency)
Concurrent Requests
--
At steady state
Minimum Cores
--
Theoretical minimum
Recommended Cores
--
With headroom
Peak Capacity
--
For traffic spikes
Utilization/Core
--
Effective Threads
--
Saturation Point
--
Max TPS (Optimal)
--
At target utilization
Max TPS (Theoretical)
--
100% utilization
Max Concurrency
--
Parallel requests
Safe TPS Limit
--
With burst headroom
Arrival Rate (λ)
--
Service Rate (μ)
--
Traffic Intensity (ρ)
--
Avg Queue Length
--
P50
--
P90
--
P95
--
P99
--
| Cores | Max TPS | Concurrency | Utilization | Status |
|---|