KUBERNETES RESOURCE FORECASTING

kube-res-foresight

Resource decisions that move with demand.

From historical metrics to future demand. From advice to controlled scaling.

What is kube-res-foresight?

Resource forecasting and scaling for a single Kubernetes cluster. Connect metrics, compare forecasts, and understand every recommendation.

Understand demand

Connect workload metrics, understand historical patterns, and forecast changes in demand.

Learn more

Review resource advice

Compare models and combine look-ahead consensus with constraints and safeguards.

Learn more

Enable bounded actions

Start with observation, then authorize admission and replica scaling while reviewing outcomes.

Learn more

CAPABILITIES

A connected workflow for every workload

Manage inputs, understand forecasts, bound actions, and keep evaluating actual outcomes.

Workload modules

One workload with its own metrics, intervals, bounds, and fallback.

Models and dynamic selection

Choose a fixed method or dynamic algorithms per module, guided by backtests and actuals.

Look-ahead decisions

Account for readiness time and combine multiple candidates into consensus.

Schedules and fallback

Define business windows and explicit behavior when inputs fail.

Memory and outage protection

Configure OOM protection and bridge outages with time-limited, scale-up-only cached decisions.

Scaling timeliness and audit

Evaluate capacity readiness against actual demand and trace advice through execution.

AI assistant and scheduled analysis

Query state with skills and schedule analyses delivered by email.

Platform activity

Review modules needing attention and applied scaling activity.

HOW IT WORKS

From metrics to action, with a feedback loop

Organize metrics and policies around a module, with clear inputs and outcomes at every stage.

Connect metrics

Templates · Data quality

Forecast & compare

Models · Rolling backtests

Recommend

Lead time · Constraints

Act & review

Admission · Scaling

Access controlScheduled tasksSafeguardsAudit

Explore the concepts

GRADUAL AUTOMATION

Start with observation. Automate at your pace.

Compare forecasts with actuals, then review resource advice. Authorize admission and scaling separately, within their respective bounds.

Safety and fallback

01ObserveObserve history and forecasts
02RecommendReview explainable recommendations
03Admission / AutoAuthorize resource and replica actions

DOCUMENTATION

Find your next step

Guides for your first workload and everyday operations.

Quick start

Connect metrics and manage your first workload.

Installation

Prepare components, settings, and startup.

Monitoring

Trace issues and review resource changes.

Start with one workload.

Understand demand with reliable metrics. Manage changes with clear bounds.

Read the documentation