# Amazon Lightsail Instance

> Lightsail bundles compute, storage, and networking into one flat monthly price per instance, accruing while the server runs. Non-production Lightsail servers therefore save directly from off-hours shutdown. ZopNight discovers instances on its 6-hour cycle, attributes bundle cost from Cost Explorer or CUR 2.0, and stops and starts them on schedule.

Source: https://zop.dev/integrations/aws/lightsail
Updated: 2026-08-19

---

Amazon Lightsail offers simplified virtual private servers with bundled compute, storage, and networking at a flat monthly price. Because the bundle price accrues while the instance runs, non-production Lightsail servers benefit directly from off-hours shutdown.

## One price, everything included

Lightsail inverts the usual AWS pricing model: instead of separate meters for compute, EBS, and data transfer, each instance carries a single bundle price covering a fixed allocation of vCPU, memory, SSD storage, and a transfer allowance. Predictability is the product. The flip side is that the bundle accrues as a unit. No per-component saving comes from using less storage or less transfer, and the only meaningful cost lever is whether the instance is running and which bundle size it runs at. Databases, container services, and load balancers in Lightsail follow the same flat-bundle logic with their own price tiers.

## Simple servers, simple schedules

ZopNight discovers Lightsail instances automatically on the 6-hour cycle, with per-resource cost attributed from Cost Explorer or CUR 2.0. Instances are schedulable: the Lightsail API stops them on the configured schedule and starts them again for working hours, which suits exactly the workloads Lightsail attracts: dev servers, internal tools, demo boxes, WordPress staging. Idle-resource recommendations catch the ones no schedule would rescue: instances whose usage signals say nobody has needed them at all.

## Where flat pricing hides waste

Lightsail's fixed price removes the cost feedback that nags people about idle EC2: the number never spikes, so the forgotten instance never draws attention. Typical leaks are demo servers from concluded sales cycles, tutorial instances from someone's learning weekend, and side-project boxes provisioned because the entry bundles cost less than lunch. Each one is trivial; together they are a steady drip. Oversized bundles are the other form: a static site on a bundle sized for a database.

## The Lightsail console view

Lightsail lives in its own console, deliberately apart from EC2, listing instances with bundle size and state per region. The instance metrics tab shows CPU and network history, enough to sort the genuinely used from the merely running before applying a schedule or a deletion.