mpc-autoscaler
MPC vs HPA
research software

Model Predictive Control for Kubernetes autoscaling, compared with HPA.

A controlled Kubernetes lab for comparing reactive HPA baselines with a short-horizon MPC controller, with current numbers, caveats, and reproduction paths up front.

60-second walkthrough Evidence Benchmark matrix Results Q&A Star / watch on GitHub Good first issues
Go 1.25 Python 3.11+ GHCR image Apache-2.0
Fast path one walkthrough before the full repository. Curated evidence claims link back to source paths. Follow progress new baselines, traces, failure cases.
01Loadstep / spike
02toy-loadGo service
03MetricsPrometheus
04PolicyHPA / MPC
05Evidencefigures / docs

Why follow this Model Predictive Control Kubernetes autoscaling project?

Star or watch the GitHub repository if you want updates on reproducible HPA vs MPC experiments, future KEDA or predictive baselines, production-like traces, failure cases, and Grafana/Prometheus evidence. The project is useful when you want an inspectable autoscaling lab, not a black-box benchmark claim.

Open the GitHub repository or browse good first issues.

Practitioner Value

Useful as an autoscaling lab, not a production replacement.

Use it to test workload knobs, metrics, HPA behavior, predictive control ideas, and evidence packaging.

Use

Inspectable workload

toy-load exposes controllable CPU, sleep, jitter, payload, error-rate, health, readiness, and Prometheus endpoints.

Compare

Policy playground

Run reactive HPA-style baselines beside MPC recommendations, then compare p95, p99, success, and replica movement.

Trust

Visible caveats

Published claims point to exact evidence paths. Missing benchmark cells stay visible instead of being implied.

Ten-Second Demo

One visual path from traffic spike to evidence.

The demo explains repository value without claiming production readiness.

Ten-second autoscaling loop demo
Traffic profile, workload metrics, HPA and MPC policies, then evidence paths and caveats.
Repository

Complete experiment loop.

Every path supports one research question: what should scale, when, and under which assumptions?

01

Control loop

Follow demand through toy-load, Prometheus metrics, HPA, MPC, and saved analysis artifacts.

Open system view

02

Reproduction tiers

Start local, move to offline simulation, then rebuild saved evidence or live cluster runs.

Run the tiers

03

Evidence policy

Charts point to committed sources and evidence roots; bulky raw artifacts stay out of Git.

View evidence

04

Project infrastructure

Release, CI, security scans, Pages, roadmap, issue templates, and GHCR image are wired as real project infrastructure.

Open project board

05

Feedback wanted

The highest-value input is specific methodology criticism: better baselines, traces, comparators, failure cases, and evidence tables.

Share reproduction feedback

06

First contribution path

Many starter tasks need only one verified documentation, link, metric, setup, or example improvement.

Browse good first issues

Evidence Preview

Evidence with source paths.

Figures summarize live runs, release automation, and reproducibility boundaries. Result claims stay tied to documented evidence paths.

MPC autoscaler architecture diagram
System boundary: traffic generator, workload, Prometheus, reactive HPA, predictive MPC, and analysis pipeline.
Spike experiment sample chart
Tuned Hybrid-SA control trace: burst detection, safety scale-up, solver status, and controlled downscale from one curated spike run.
release

v0.1.0 shipped

Checksums, binaries, Helm package, GHCR image, SBOM, and provenance are published.

Open release