---
title: "Building cloud-native PACS on AWS"
description: "Multi-hospital networks face rising imaging volumes, siloed on-premises PACS, radiologist routing bottlenecks, and storage that must expand for 7 to 10 years of retention. The proposed pattern uses a hub-and-spoke hybrid architecture: local PACS instances preserve LAN-speed reading, while a centralized AWS archive provides cross-facility discovery, disaster recovery, and lifecycle-managed storage. Modalities send DICOM through C-STORE to local VNA servers, which normalize and compress objects; asynchronous replication sends images to Amazon S3, while Amazon Aurora PostgreSQL stores metadata and routing information. The AWS deployment spans two Availability Zones, places PACS services behind Network Load Balancers, and includes CloudFront and AWS WAF. Storage guidance distinguishes S3 Intelligent-Tiering, Glacier Instant Retrieval, and Glacier Deep Archive according to access frequency and retention needs."
---

# Building cloud-native PACS on AWS

[AWS](https://yomu.fyi/company/aws) · ManojKumar MV · Sep 17, 2026

**Type:** Explainer

## Summary

Multi-hospital networks face rising imaging volumes, siloed on-premises PACS, radiologist routing bottlenecks, and storage that must expand for 7 to 10 years of retention. The proposed pattern uses a hub-and-spoke hybrid architecture: local PACS instances preserve LAN-speed reading, while a centralized AWS archive provides cross-facility discovery, disaster recovery, and lifecycle-managed storage. Modalities send DICOM through C-STORE to local VNA servers, which normalize and compress objects; asynchronous replication sends images to Amazon S3, while Amazon Aurora PostgreSQL stores metadata and routing information. The AWS deployment spans two Availability Zones, places PACS services behind Network Load Balancers, and includes CloudFront and AWS WAF. Storage guidance distinguishes S3 Intelligent-Tiering, Glacier Instant Retrieval, and Glacier Deep Archive according to access frequency and retention needs.

## Context

The post is aimed at healthcare IT architects and solutions architects familiar with DICOM workflows. It addresses the limits of independently operated on-premises PACS across multi-hospital networks, including storage growth, data silos, routing bottlenecks, and long-term retention.

## Approach / What changed

Use a hub-and-spoke hybrid deployment in which each hospital retains local PACS components for immediate LAN-based imaging workflows while connecting to a centralized AWS archive through AWS Direct Connect or AWS Site-to-Site VPN. Local VNA servers ingest, normalize, compress, and asynchronously replicate DICOM objects to Amazon S3. Centralized PACS services use Amazon Aurora PostgreSQL for metadata, run across two Availability Zones behind Network Load Balancers, and apply storage lifecycle policies based on access patterns.

## Takeaways

- Local PACS instances keep newly acquired studies available at LAN speed, while asynchronous synchronization to Amazon S3 prevents cloud replication from blocking clinical workflow.
- A common PACS vendor and shared metadata database allow radiologists to discover and retrieve studies created at any hospital in the network.
- Cloud-only versus hybrid deployment depends primarily on redundant connectivity and vendor performance; imaging volume alone does not determine the choice.

## FAQ

### How does the hybrid design keep clinical reading available during cloud synchronization?

Imaging devices send studies to the local VNA over the hospital LAN, and studies are immediately available for radiologist reading there. Replication to Amazon S3 occurs asynchronously, so the clinical workflow is not blocked by synchronization.

### How are different storage tiers matched to imaging access patterns?

S3 Intelligent-Tiering is intended for unpredictable access, Glacier Instant Retrieval for studies accessed occasionally such as older comparative exams, and S3 Glacier Deep Archive for studies older than five years requiring long-term retention. Deep Archive retrieval takes 12 to 48 hours.

### When is cloud-only PACS a stronger fit than hybrid PACS?

Cloud-only PACS is a stronger fit when a facility has redundant, reliable, high-bandwidth connectivity and a vendor solution that performs as well or better in the cloud. Hybrid PACS fits sites with limited connectivity or solutions optimized for local caching.

**Tags:** [Architecture](https://yomu.fyi/topic/architecture), [AWS](https://yomu.fyi/topic/aws), [DICOM](https://yomu.fyi/topic/dicom), [Healthcare](https://yomu.fyi/topic/healthcare), [S3](https://yomu.fyi/topic/s3)

- Source: [AWS](https://aws.amazon.com/blogs/architecture/building-cloud-native-pacs-on-aws)
- Source URL: https://aws.amazon.com/blogs/architecture/building-cloud-native-pacs-on-aws
- Ingested by Yomu: 2026-09-17T15:23:08.703Z

[Read original post](https://aws.amazon.com/blogs/architecture/building-cloud-native-pacs-on-aws)