ADAPT: A Self-Calibrating Proactive Autoscaler for Container Orchestration

TL;DR

ADAPT introduces an online estimator and a dynamic planning approach for proactive autoscaling in container orchestration, significantly reducing SLA violations across diverse workloads.

Contribution

It presents a novel self-calibrating autoscaler that adapts its lookahead based on measured provisioning delays using an EWMA estimator and MPC, improving scaling accuracy.

Findings

MPC+LSTM achieves below 5% SLA violation on all workloads.

Compared to reactive HPA, MPC+LSTM reduces violations from 7-19%.

MPC+Prophet can have violations up to 28.7% on bimodal traffic.

Abstract

Proactive autoscaling for containerized workloads depends on knowing the provisioning delay, i.e., the time between a scaling decision and the moment new capacity is ready to serve traffic. In practice, this cold-start duration can vary substantially across environments and even across consecutive scale-out events. We present ADAPT (Adaptive Duration Approximation for Predictive Timing), an online EWMA estimator that tracks coldstart duration at runtime. ADAPT feeds a dynamic planning horizon, FH-OPT, into a Model Predictive Controller (MPC) that optimizes replica counts over a rolling window. Together, these components form a closed-loop proactive autoscaling design that adapts its lookahead based on measured provisioning delay. Evaluated across three policies (MPC+LSTM, MPC+Prophet, HPA) and six workload archetypes with five random seeds, MPC+LSTM achieves below 5% SLA violation on all workloads, compared with 7-19% for reactive HPA and up to 28.7% for MPC+Prophet on bimodal traffic.