Openenv Autoscaling Control Center
Openenv Autoscaling Control Center is a packaged OpenEnv environment for training and evaluating agents on cloud-service autoscaling and service stability tasks.
Core trick
Simulates a Kubernetes-like service with startup-delayed scale-up actions so policies must anticipate traffic, scored on latency, queue depth, and cost.
Tags
Description
Openenv Autoscaling Control Center is a packaged OpenEnv environment for training and evaluating agents on cloud-service autoscaling and service stability tasks. The environment simulates a Kubernetes-like service under changing traffic conditions and exposes an action space for scaling decisions. We built a formal OpenEnv package, a heuristic baseline and RL training scaffolding, and a demo UI.
The environment is designed to test whether agents can make good operational decisions under dynamic load. It includes realistic metrics such as incoming requests per second, ready pods, queue depth, latency, and reward. I also added richer scenario families such as traffic spikes, bad deploys, and dependency slowdowns in our advanced branch to move toward incident-response style control.
README excerpt
# OpenEnv Autoscaling Environment ## Project Overview This project implements a Kubernetes-like autoscaling simulator and packages it as a formal OpenEnv environment. It includes: - A deterministic simulator for one autoscaled service under changing load - A formal OpenEnv package at `envs/autoscale_env` - Baseline heuristic policy + evaluation scripts - SFT and Unsloth+GRPO training scaffolding - Colab notebook and script for packaged-environment RL training ## Why This Environment Is Interesting - **Delayed scaling dynamics**: scale-up actions have startup delays, so action quality depends on anticipation. - **Real tradeoffs**: reward balances latency SLO violations, queue buildup, error rate, cost, and scaling flaps. - **Policy-learning friendly**: deterministic traces and explicit action space make it suitable for imitation + RL experiments. - **Hackathon practical**: includes end-to-end generation, baseline, packaging, smoke tests, and Colab training flow. ## Formal OpenEnv Package Status The primary environment is now a formal OpenEnv package under `envs/autoscale_env` with: - `openenv.yaml` - `pyproject.toml` - packaged server/client/models/environment modules - validated deployment modes (`openenv_serve`, `uv_run`, `python_module`) Validation command: ```bash cd envs/autoscale_env openenv validate --verbose ``` ## Repo Structure - `envs/autoscale_env/` - primary packaged OpenEnv environment (submission path) - `simulator.py` - simulator source of truth (st