{"id":197,"date":"2026-10-12T18:21:24","date_gmt":"2026-10-12T18:21:24","guid":{"rendered":"http:\/\/www.scmgalaxy.com\/tutorials\/2010\/03\/10\/running-java-ee-applications-on-amazon-ec2\/"},"modified":"2026-10-12T18:21:24","modified_gmt":"2026-10-12T18:21:24","slug":"running-java-ee-applications-on-amazon-ec2","status":"publish","type":"post","link":"https:\/\/www.devopsschool.com\/blog\/running-java-ee-applications-on-amazon-ec2\/","title":{"rendered":"Deploying Java EE on Amazon EC2 with EC2Deploy: Workflow, Script, and Modern Lessons"},"content":{"rendered":"\n

Originally published on 2010-03-10; fully revised and updated.<\/em><\/p>\n\n\n\n

This article explains a classic pattern for running Java EE applications on Amazon EC2<\/a> using a Groovy framework called EC2Deploy. If you maintain legacy deployment automation or want to understand early cloud-era deployment design, this walkthrough shows the end-to-end flow: infrastructure launch, database and app server setup, WAR deployment, load testing, and environment teardown.<\/p>\n\n\n\n

Why this approach mattered for Java EE teams<\/h2>\n\n\n\n

The original post addressed a real pain point: performance and environment testing required hardware that was expensive and slow to obtain. EC2 made on-demand servers possible, and EC2Deploy wrapped that capability into a developer-friendly API for multi-node test environments.<\/p>\n\n\n\n

The source article also referenced then-current EC2 pricing and server-count limits, plus selectable server types<\/a>. Treat those figures as historical context, not current AWS pricing or quotas.<\/p>\n\n\n\n

EC2Deploy workflow and architecture<\/h2>\n\n\n\n

At a high level, EC2Deploy reads AWS properties, builds a cluster specification, starts EC2 instances, configures MySQL\/Tomcat\/Apache, deploys the application, runs JMeter tests, and then stops the instances.<\/p>\n\n\n\n

\nflowchart TD\n    A[Groovy test script] --> B[Load aws.properties]\n    B --> C[Build ClusterSpec]\n    C --> D[Call cluster.start]\n    D --> E[Launch EC2 instances]\n    E --> F[Init MySQL master and slave]\n    F --> G[Upload DML and WAR to S3 cache]\n    G --> H[Deploy app on Tomcat]\n    H --> I[Configure Apache load balancer]\n    I --> J[Run cluster.loadTest with JMeter]\n    J --> K[Collect CPU latency throughput report]\n    K --> L[Call cluster.stop]\n<\/pre>\n\n\n\n
Core API objects and methods<\/h2>\n\n\n\n
The original code uses a small set of objects and fluent methods to describe both infrastructure and application behavior.<\/p>\n\n\n\n
Object or method<\/th>Role in the workflow<\/th><\/tr><\/thead>
AWSProperties(“\/path\/to\/aws.properties”)<\/td>Loads AWS credentials and AMI-related properties.<\/td><\/tr>
new EC2(awsProperties)<\/td>Creates the EC2 client wrapper used by EC2Deploy.<\/td><\/tr>
ClusterSpec.schema(…)<\/td>Defines DB schema name, user\/password map, and DML scripts.<\/td><\/tr>
ClusterSpec.slaves(1)<\/td>Specifies the number of MySQL slave nodes.<\/td><\/tr>
ClusterSpec.tomcats(1)<\/td>Specifies the number of Tomcat nodes.<\/td><\/tr>
ClusterSpec.webApp(explodedWar, “ptrack”)<\/td>Declares exploded WAR path and context root.<\/td><\/tr>
ClusterSpec.catalinaOptsBuilder(…)<\/td>Builds Tomcat JVM and app properties, including DB host wiring.<\/td><\/tr>
cluster.start()<\/td>Provisions instances and performs configuration\/deployment steps.<\/td><\/tr>
cluster.loadTest(“…SimpleTest.jmx”, [1,2])<\/td>Runs JMeter tests at multiple thread counts.<\/td><\/tr>
cluster.stop()<\/td>Stops the EC2 instances when testing is complete.<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
Annotated Groovy example<\/h2>\n\n\n\n
The following script is a cleaned-up version of the original example, preserving the same structure and intent.<\/p>\n\n\n
class<\/span> ClusterTest<\/span> extends<\/span> GroovyTestCase<\/span> <\/span>{\n  void<\/span> testSomething() {\n    AWSProperties awsProperties = new<\/span> AWSProperties(\"\/path\/to\/aws.properties\"<\/span>)\n    def ec2 = new<\/span> EC2(awsProperties)\n\n    def explodedWar = \"\/path\/to\/projecttrack\/webapp\/target\/ptrack\"<\/span>\n\n    ClusterSpec clusterSpec = new<\/span> ClusterSpec()\n      .schema(\"ptrack\"<\/span>, [\"ptrack\"<\/span>: \"ptrack\"<\/span>], [\n        \"src\/test\/resources\/testdml1.sql\"<\/span>,\n        \"src\/test\/resources\/testdml2.sql\"<\/span>\n      ])\n      .slaves(1<\/span>)\n      .tomcats(1<\/span>)\n      .webApp(explodedWar, \"ptrack\"<\/span>)\n      .catalinaOptsBuilder({ builder, databasePrivateDnsName ->\n        builder.arg(\"-Xmx500m\"<\/span>)\n        builder.prop(\"com.sun.management.jmxremote\"<\/span>)\n        builder.prop(\"com.sun.management.jmxremote.port\"<\/span>, 8091<\/span>)\n        builder.prop(\"com.sun.management.jmxremote.authenticate\"<\/span>, false<\/span>)\n        builder.prop(\"com.sun.management.jmxremote.ssl\"<\/span>, false<\/span>)\n        builder.prop(\"ptrack.application.environment\"<\/span>, \"ec2\"<\/span>)\n        builder.prop(\"log4j.configuration\"<\/span>, \"log4j-minimal.properties\"<\/span>)\n        builder.prop(\"jdbc.db.server\"<\/span>, databasePrivateDnsName)\n      })\n\n    SimpleCluster cluster = new<\/span> SimpleCluster(ec2, clusterSpec)\n    cluster.start()\n    cluster.loadTest(\"\/path\/to\/functionalTests\/jmeter\/SimpleTest.jmx\"<\/span>, [1<\/span>, 2<\/span>])\n    cluster.stop()\n  }\n}<\/code><\/span>Code language:<\/span> JavaScript<\/span> (<\/span>javascript<\/span>)<\/span><\/small><\/pre>\n\n\n
What happens during cluster.start()<\/h2>\n\n\n\n