Skip to main content
This lesson explains AWS Application Discovery Service — what it collects, how it works, and how it supports migration planning and execution. Application Discovery Service gathers on‑premises inventory, utilization metrics, and dependency mappings so AWS Migration Hub and other migration tools can plan, prioritize, and track migrations. By revealing relationships between applications, servers, processes, and network communications, it helps teams design move groups, avoid hidden dependencies, and right‑size target resources.
A slide titled "Dependency Mapping" showing a corporate data center diagram with a server and two databases connected by arrows. To the right are icons representing an application discovery/service and a network discovery graphic.

​
Key capabilities

​
Discovery methods — agent-based vs agentless

Choose the collection method that best matches your environment, visibility needs, and operational constraints. The service supports two primary approaches: agent-based discovery (per‑server visibility) and agentless discovery (centralized collection for VMware).
A diagram titled "Agent Discovery" showing a corporate data center with a server and several agents that connect over TLS to an Application Discovery Service. Icons on the right indicate collected items like system specifications, performance data, and network connections.
Agent-based discovery
  • Deploy a lightweight Application Discovery agent on each server to capture system and application‑level data.
  • Agents collect running processes, performance metrics, open ports and network connections, and configuration details.
  • Data is transmitted securely to Application Discovery Service over TLS at regular intervals.
  • Best when you need process‑level visibility, application traffic detail, and per‑server diagnostics.
A diagram titled "Agentless Discovery" showing a data center with multiple VMs and a VMware vCenter Server connecting (via an agent) to an Agentless Collector. To the right are icons listing collected items like CPU, RAM, database metadata, and utilization metrics.
Agentless discovery
  • Deploys an agentless collector as a virtual appliance in your VMware vCenter environment.
  • The collector queries vCenter APIs to capture VM/host metadata, utilization metrics, and database metadata without installing agents on each VM.
  • Lower per‑server maintenance and useful for large VMware estates where per‑process detail is not required.
Use agent-based discovery when you need process-level visibility and fine-grained network communication data. Use agentless discovery when you prefer a centralized, lower-maintenance approach for VMware environments and do not require per-process information.
Table — quick comparison

​
Data flow and storage

Regardless of method, collected discovery data flows securely to the Application Discovery Service and is exported for storage and analysis:
  • Agents or collectors send discovery packets over TLS at regular intervals.
  • Raw and processed discovery output can be exported to Amazon S3 for long‑term retention.
  • Use Amazon Athena to query S3‑stored discovery data and generate custom reports or dashboards.
A simple architecture diagram showing servers and discovery agents in a data center sending data to an Application Discovery Service, which stores output in Amazon S3 and is queried by Amazon Athena.

​
Integrations and downstream tools

Application Discovery Service feeds other migration services and analytics tools to accelerate planning and execution:
  • AWS Migration Hub: Import discovery inventories and dependencies to build migration plans and track progress.
  • AWS Database Migration Service (DMS): Use discovery data to identify database sources and plan migrations.
  • Migration Evaluator (formerly TSO Logic): Analyze utilization and estimate migration costs and ROI.
  • Amazon S3 + Amazon Athena: Store and query raw discovery data for custom analytics and reporting.

​
Common use cases

  • Inventory discovery: Build a reliable, searchable inventory of on‑premises servers and databases to accelerate migration assessments.
  • Dependency mapping: Discover communication patterns and hidden dependencies to create safer move groups and reduce application downtime risk.
  • Right‑sizing and capacity planning: Use historical utilization metrics to recommend appropriate Amazon EC2 instance types and estimate target infrastructure costs.
Application Discovery Service supplies the foundational insights required to plan migrations confidently and to fuel downstream AWS migration tools for execution and monitoring.

Watch Video