Scheduler Capacity Calculator
Estimate how many pods per second the Kubernetes scheduler can place on a cluster.
About Scheduler Capacity Calculator
The Kubernetes scheduler's throughput is bounded by scheduling latency and parallelism: throughput = (1000/latencyMs) × threads. The percentageOfNodesToScore setting limits how many nodes are evaluated per pod, trading placement quality for speed in large clusters.
Formula
Throughput
podsPerSec = (1000 / schedulingLatencyMs) × threads
- schedulingLatencyMs
- Average time to schedule one pod
- threads
- Parallel scheduling workers
Worked example
With the default values below, Scheduler Capacity Calculator returns Scheduling Throughput of 400.0. Change any input above and every figure here is recalculated in your browser.
Inputs used
- Nodes in Clusteraccepts 1–100,000 nodes
- 50 nodes
- Avg Scheduling Latencyaccepts 0.1–10,000 ms
- 5 ms
- Scheduler Parallelismaccepts 1–32 threads
- 2 threads
- Nodes to Score %accepts 1–100 %
- 50 %
Results
- Scheduling Throughput
- 400.0 pods/s
- Pods per Minute
- 24,000 pods/min
- Nodes Scored per Pod
- 25 nodes
- Per-Thread Throughput
- 200.0 pods/s
Step by step
- 1Nodes scored per pod= 25
- 2Pods/sec per thread= 200.0
- 3Total scheduling throughput= 400.0 pods/s
Assumptions and limits
- The result is derived only from the 4 inputs listed above. Anything not among them is not part of the calculation.
- In particular, this calculator does not take inflation, tax, fees or charges or compounding frequency into account.
- Figures are rounded for display to 1 decimal place, so totals built from several rounded parts can differ slightly from the unrounded arithmetic.
How these calculators are built and tested is described on the Methodology page. If a figure looks wrong, please report it.
Frequently Asked Questions
What increases scheduling latency?
Complex affinity/anti-affinity rules, many plugins, large numbers of nodes to score, and resource pressure on the control plane all increase per-pod scheduling time.
When does scheduler throughput become a bottleneck?
In clusters with rapid scaling (thousands of pods launching simultaneously, e.g. batch jobs or cluster-wide rolling updates), the scheduler can become a bottleneck if throughput is lower than the pod creation rate.