OpenMRS on AWS - Hardened EMR Platform, CloudWatch & SSM Ready

Deploy a production ready OpenMRS electronic medical record system on AWS in minutes. Built for clinics, NGOs, global health programs and research teams who need the world's most widely deployed open source EMR without days of Java, Tomcat and database assembly. This image ships the OpenMRS 3 platform with its modern frontend on a CIS aligned, hardened base OS.

Why OpenMRS, and why this image

OpenMRS powers patient records for health systems in more than forty countries and underpins some of the largest HIV, TB and primary care programs in global health. It is modular, standards aware (FHIR support via module) and backed by a large implementer community. What it is not is simple to stand up: the platform, frontend, Tomcat, Java and database each have version interdependencies that consume implementer time better spent on clinical configuration. This image arrives assembled, secured and documented, so your first day is spent on forms and workflows, not dependency resolution.

What you get out of the box

• OpenMRS 3 platform and frontend at current stable versions, with the version matrix (Java, Tomcat, MariaDB) resolved and recorded in the image manifest
• MariaDB backend pre configured with the OpenMRS schema, UTF8MB4 throughout
• Automated daily database and application data backup scripting with retention configuration

Common use cases

• Clinics and hospital departments: patient registration, visits, encounters, orders and reporting
• NGO and global health programs: the OpenMRS heartland; multi site deployments with offline aware workflows via community modules
• Research and academic programs: structured clinical data capture with export paths
• Training environments: safe, disposable EMR instances for clinical informatics education
• Evaluation: assess OpenMRS against commercial EMRs on your own infrastructure with no licence cost

Compliance: read this before storing patient data

An AMI cannot make a deployment compliant; operations do. This image gives you a hardened starting point, and the documentation is explicit about the remaining shared responsibilities: Consider using encrypted volumes, network isolation, access control, audit trails, backup handling. Place instances in private subnets where clinical workflows allow, restrict the web tier to your networks, and never expose the database port.

Usage notes

• Attach an IAM instance role with the AWS-managed policies CloudWatchAgentServerPolicy and AmazonSSMManagedInstanceCore to activate CloudWatch and Session Manager. No AWS credentials are stored on the image.
• Allow about five minutes after launch. Clinical app: http:/// ; admin/legacy UI: http:///openmrs (same login). Credentials: /opt/openmrs/default-openmrs-credentials.txt; database passwords in /opt/openmrs/.env (root-only). Manage the stack with cd /opt/openmrs && docker compose ps|logs -f|restart.
• Ports: 22/tcp (SSH), 80/tcp (HTTP). The stack serves plain HTTP - put TLS in front (load balancer or nginx + certbot) for production.
• CloudWatch agent config: /opt/aws/amazon-cloudwatch-agent/etc/amazon-cloudwatch-agent.d/openmrs.json

Quick Start:

• Allow ~5 minutes after launch for first-boot initialization.
• Clinical app: http:/// - Admin/legacy UI: http:///openmrs (same login).
• Credentials were generated on first boot: SSH in as 'ubuntu' and read /opt/openmrs/default-openmrs-credentials.txt.
• For production, put TLS in front (load balancer or nginx + certbot).
• Monitoring: attach an IAM instance role with the AWS-managed policies CloudWatchAgentServerPolicy and AmazonSSMManagedInstanceCore to enable CloudWatch logs/metrics (namespace CloudSOE/openmrs) and Session Manager. No AWS credentials are stored on the image.

Sizing guidance

Evaluation and training: t3.large (Java needs memory; smaller instances will disappoint). Single facility production: m7i.large with encrypted gp3 storage. Multi site or heavy reporting: m7i.xlarge and consider RDS for MariaDB, documented in the architecture guide.

OpenMRS® is a registered trademark of OpenMRS Inc. This product is independently packaged and maintained and is not affiliated with or endorsed by OpenMRS Inc. or the OpenMRS community.