Available tool
Database Connection Contention and Recovery Lab
Model bounded connection pools, lock waits, retries, replica lag, and failover ambiguity in a deterministic browser-only timeline.
Open toolReliability Labs
Kubernetes Rollout Capacity Lab
Model readiness, surge, unavailable replicas, drain time, rollout stalls, and version capacity with a deterministic browser-only timeline.
Deterministic browser-only rollout model
Kubernetes Rollout Capacity Lab
Compare readiness, surge, unavailability, termination, and modeled serving capacity without contacting or changing a cluster.
This is not a benchmark, capacity calculator, Kubernetes dry-run, or production verification. Steps and capacity are explicit teaching inputs, not measured time or throughput.
Six teaching scenarios
Load a fixed rollout boundary
Deterministic result
Rollout timeline and modeled capacity
| Step | Old Ready | New starting | New Ready | Terminating | Unavailable | Capacity | Risk | Action |
|---|
Production evidence to inspect next
How to use it
Choose one of six scenarios or enter a bounded teaching configuration. The model advances old Ready, new starting, new Ready, and terminating Pods through discrete steps, then reports rollout progress and modeled serving capacity.
The inputs are assumptions and policy values, not observations from a cluster. A timeline step is not a second. A Pod capacity value is not a benchmark. The tool does not generate YAML or production recommendations.
Read Kubernetes Rollout Capacity Lab: Readiness, Surge, and Drain for every preset and its evidence boundary. Continue with Safe Backend Releases and Production Kubernetes Incident Troubleshooting.
This lab is not a benchmark, capacity calculator, Kubernetes dry-run, or production verification. Nothing leaves this browser.
Related tools
Keep the workflow moving
These local-first tools often pair well with Kubernetes Rollout Capacity Lab.
Available tool
Java Concurrency Budget Lab
Model fixed pools, virtual threads, resource guards, deadlines, and cooperative cancellation with a deterministic browser-only timeline.
Open toolAvailable tool
Message Delivery Semantics Lab
Simulate ACK timing, crashes, bounded retries, redelivery, duplicate effects, and stable idempotency entirely in your browser.
Open toolFAQ
Kubernetes Rollout Capacity Lab questions
Does this lab connect to or modify a Kubernetes cluster?
No. It is a deterministic browser-only state model. It makes no network request, runs no kubectl command, and reads no cluster, registry, cloud, or account state.
Can this lab size production replicas or resource limits?
No. Capacity and timeline steps are teaching inputs rather than measurements. Production policy needs representative workload, startup, resource, dependency, and user-outcome evidence.
How are percentage surge and unavailable values resolved?
The model demonstrates Deployment-style rounding: percentage maxSurge rounds up, while percentage maxUnavailable rounds down for the selected desired replica count.
What does overload risk mean in the output?
It only means the entered request rate exceeds ready modeled Pod capacity in that step. It does not predict latency, errors, scheduling, OOM, or business outcomes.