Step 2. Migrate legacy JSP system from Windows to AWS Linux — Migrate Product Creation System

I have a client with a legacy Java/JSP system built in the mid 2000's that has remained virtually unchanged. Up to now, they have been self-hosted on a Windows 2003 platform. They are looking to update the system to operate on AWS Linux, and migrate it. Currently, the system is not working in production, and the client needs to use it, so we would like to execute this project as soon as feasible. This system is built on JAVA/JSP with Tomcat application server, Apache web server, and MySQL. There are several small, supporting applications that run as services, as well as a number of hard-coded references to the file system, and a few other dependencies on the Windows OS. The system has two main aspects: the front end, public (and staff) facing web site to interact with the system, and a processing system that creates information products. It was built as a distributed system with the front end on one instance, and one or more processing node instances to create products. The client now only services a handful of customers with (usually) only a single individual using the system, so they would like to maintain a single instance with the option of adding processing nodes if necessary. The front end instance includes the code for processing products, and processing nodes have the JSP pages and servlets removed. Step 2. Migrate Product Creation System This step is dependent on the following job: Step 1. Migrate legacy JSP system from Windows to AWS Linux — Migrate Front End https://www.freelancer.com/projects/software-architecture/step-migrate-legacy-jsp-system/ Once we have a stable development instance for the front end, we can focus on getting product creation working. This system creates geospatial imagery and mapping digital products, and generates as part of those products, a PDF report. There are three main dependencies on the Windows OS that we will need to address (we can adjust scope if we discover others). 1. File system references — we'll need to update the system to read/store on S3 2. PDF Creation — currently to create a PDF, the system exports the report as HTML, instantiates Internet Explorer on the processing node, and "prints" the web page to a PDF. We would like to do away with this awkward solution, and build PDFs using a more elegant solution that is OS agnostic. 3. Executable zip — an earlier requirement was that the final product should be stored as an executable, self-extracting zip. We would like to eliminate this step, and just store the product as a zip. We expect there may be other issues to tackle, and we can handle them as they come. The objective and requirement is to arrange it so products are created reliably while hosting the entire system on a single, linux instance.