Skip to content

Introduction

Lambda Managed Instances (LMI) runs your function on dedicated EC2
Elastic Compute Cloud — the AWS virtual machine service that underpins Lambda Managed Instances.
capacity in your own VPC
Virtual Private Cloud — an isolated, private network in AWS where your Lambda Managed Instances and related resources run.
instead of Shared Lambda infrastructure (the multi-tenant Lambda
AWS Lambda — a serverless compute service that runs code in response to events without provisioning servers.
fleet). You still use the Lambda programming model; compute is EC2 in your account, wired through a capacity provider. Isolation and tenancy differ from default Lambda — see Tenancy & isolation.

BenefitWhat it means
FlexibilityCurrent-gen CPUs, Graviton, network-optimised options — matched to workload
ControlYour VPC, subnets, security groups — not another customer’s fleet
PerformanceConfigurable memory / vCPU ratios and bandwidth
Cost leversEC2 Savings Plans and RIs apply to underlying compute (management fee is separate)

The figure above shows the operational sequence for LMI:

  1. Provision placement — define a capacity provider (VPC, subnets, security groups, instance requirements, scaling limits).
  2. Wire the function — attach capacity_provider_config on the Lambda to that provider.
  3. Publish — a new version remains inactive until managed instances and execution environments are ready; it then becomes ACTIVE. Another function on the same provider may consume existing spare capacity instead of adding instances.

Fully managed fleet: you configure VPC, provider, and function; Lambda operates managed instances (lifecycle, patching, routing, and fleet scaling policy on that surface are not yours to run like normal EC2).

How this site is organised

Content is released in batches. The sidebar lists published pages only (draft pages are omitted from production builds).

AreaRole
Getting startedThis page — theory orientation, no procedural steps
LMI fundamentalsShort articles: placement through publishing, networking, monitoring, quotas
Decision guide & referenceEconomics, integrations, constraints, patterns, Terraform sketches
WalkthroughOverview — hands-on entry; Prerequisites and future IAM / build / demo pages
Demo (under Walkthrough when enabled)End-to-end example when that batch publishes

The following articles each cover one concern—architecture choices, limits, and behaviour you need before you build. Read in any order. For quotas, API fields, and region-specific behaviour, use Lambda Managed Instances as the product reference.