solutions / autoscaling
Scales your infra when it needs to.
Scaling as a Service
- Keep your business running through high volume requests.
- Scale down again during lower demand, and stop paying for idle.
- Integrated with inexpensive, popular public clouds.
- Driven by the metrics you already collect, not by guesswork.
How it works
scalr watches the metrics you already collect — requests per second, queue
depth, CPU saturation, anything you can express as a number — and adjusts the
size of your instance pool to match. When demand rises it provisions from your prefered cloud provider’s API. When demand falls it drains and removes the surplus.
What you need
A cloud account we support, a metric source, and a machine image that boots into your application. That is the whole list. There is no agent to install on your instances and no control plane of ours sitting in the request path.
Where it runs
Vultr, Hetzner Cloud, Exoscale, DigitalOcean, cloudscale.ch, Proxmox and Apache CloudStack today. Adding a provider is a contained piece of work — if you run somewhere else, ask us.