Instance disk performance (IOPS & throughput)

Every Lite instance gets disk performance derived from a single fixed policy — there is no per-plan IOPS column and no separate performance tier. The provisioned disk size is the only input, and the same formula is applied to every plan family at provisioning time.

This is intentional. We do not overprovision the underlying hardware to let a few workloads spike at everyone else’s expense — our goal is to deliver predictable, uniform disk performance to every customer. The numbers below are what you actually get on a sustained basis, not best-case marketing figures or burst windows that quietly disappear under load.

The policy at a glance

• Base capacity: 100 GB included at the baseline.
• Base IOPS: 8,000 — the floor for every instance, regardless of size.
• IOPS scaling: +16 IOPS per GB above the 100 GB baseline.
• Maximum IOPS: 48,000 — hard cap once scaling exceeds this point.
• Base throughput: 400 MB/s — the floor for every instance.
• Throughput scaling: +4 MB/s per GB above the 100 GB baseline.
• Maximum throughput: 1,500 MB/s — hard cap.
• Burst: not available. The numbers above are sustained, not peak.

How it scales by disk size

IOPS grows linearly with the disk size you provision until it hits the 48,000 cap at 2,600 GB. Below 100 GB you stay at the 8,000 IOPS floor.

• 100 GB → 8,000 IOPS · 400 MB/s
• 200 GB → 9,600 IOPS · 800 MB/s
• 500 GB → 14,400 IOPS · 1,500 MB/s (throughput capped)
• 1,000 GB → 22,400 IOPS · 1,500 MB/s
• 2,600 GB and above → 48,000 IOPS · 1,500 MB/s (both capped)

What this means for plan selection

Because the policy is identical across plan families, the only lever to dial up disk performance is to provision a larger disk size at deploy time. CPU, RAM, and network bandwidth still differ between families — but disk IOPS and throughput follow the same curve everywhere.