Spot Autoscaling

Status: Coming soon

Spot Autoscaling will automate the use of cloud provider spot/preemptible instances for fault-tolerant workloads, capturing the deepest available discounts for interruptible compute.

What it will do

Spot instances (AWS), preemptible VMs (GCP), and spot VMs (Azure) offer discounts of 60% to 90% compared to on-demand pricing. The tradeoff is that the cloud provider can reclaim these instances at any time when capacity is needed elsewhere.

Pump's Spot Autoscaling will manage this tradeoff by:

1. Identifying eligible workloads. Analyzing your running workloads to determine which can tolerate interruption (batch processing, CI/CD, stateless web tiers, data pipelines).

2. Managing spot lifecycle. Handling instance bidding, interruption notices, and automatic replacement.

3. Diversifying across pools. Spreading spot usage across multiple instance types and availability zones to reduce interruption risk.

How it relates to other Pump Save mechanisms

Mechanism	Layer
Commitment Planner	Pricing: discounts on committed, predictable usage
Infrastructure Recommendations	Sizing: right-size or eliminate underused resources
Kubernetes Auto-Scaling	Pod sizing: reduce compute per pod
Spot Autoscaling	Instance type: shift eligible workloads to the cheapest available compute

Spot savings apply on top of right-sizing. A workload that has been right-sized and moved to spot instances captures savings at both layers.

When it will be available

Spot Autoscaling is currently in development. If you're interested in early access, contact your account team or reach out to [e-mail].