Three keys for moving federal ERP to the cloud

Agencies across the federal government are embracing a “cloud first” approach to modernization, but not every type of system is easily adapted to a cloud platform. Enterprise resource planning (ERP) systems are among the most complex legacy platforms that agencies need to modernize. Federal ERP systems handle federal financial management, and are the backbone of agency operations.

They are critical to running the business of government, essential in budgeting for, procuring and paying for goods and services. Given the sheer scale of transactions the government undertakes to run its operations, modernizing ERP systems is often a multi-year project. 

Understanding ERP

There are no shortcuts for ERP cloud migration. A successful move requires an intimate understanding of your agency’s ERP processes and system architectures. Based on CGI’s experience, we suggest agencies keep these three things in mind when considering a cloud migration: 

Understanding the ERP architecture is critical. Data- and transaction-rich, ERP architectures are by nature highly complex. Within traditional hosting models, various virtual machine (VM) types are optimized for the ERP, based upon the type of load. When designing your to-be cloud architecture for an ERP system, you must understand what types of workloads the VMs will support, and then design and provision each VM to fulfill its specific task. Agencies looking to optimize their administrative and operational capabilities through a more composable ERP strategy may look to a hybrid cloud model to achieve optimum performance. 

Some workloads, such as database management, will require premium storage to support input/output demands and perform with minimal latency. Memory-optimized virtual machines are well suited for online transaction processing components of the ERP. Your cloud architect must have a deep understanding of what transaction types are supported by each element of the architecture to identify the optimum cloud services to meet those needs – or improve performance – moving forward. 

By nature, ERP systems integrate with other agency systems necessary to perform the business of government. In most cases, the migration of the ERP system to the cloud will be independent of any modernization projects for these integration points. An ERP-knowledgeable cloud architect who understands the integration points from technology, data and mission delivery perspectives can assist in architecting a cost-effective, high-performing model. 

Envision and architect for your data future. Cloud architects must understand where data resides (including personally identifiable information) within your ERP’s complex data structure. Detailed knowledge of ERP data architecture supports an approach that optimizes use of native encryption capabilities, particularly to support sensitive data. At the same time, strong role-based access controls within the cloud enable a more finely tuned access paradigm. 

Most large federal ERP systems continue to operate on legacy relational database management systems. Others may leverage in-memory databases within their architecture, where partitioning is key to meeting performance requirements cost-effectively. When considering a future cloud-based state, agencies should consider architecture elements such as containerization alongside platform-as-a-service offerings to meet high availability needs. 

Working with our clients, CGI has migrated many federal ERP systems’ existing databases to a variety of cloud platforms, including Amazon Web Services and Microsoft Azure, enabling agencies to:

• Reduce database administration overhead, resulting in approximately 30% cost savings
• Enhance availability by leveraging cloud-native high availability capabilities
• Improve security with cloud-native security configurations

Optimize your approach to disaster recovery. The migration of an ERP system to the cloud provides an excellent opportunity to refresh the agency’s approach to disaster recovery. Cloud architectures provide a host of disaster recovery options that are readily accessible and easier to implement than traditional hosting models. The cloud’s inherent elasticity enables agencies to evolve to a more robust approach cost-effectively. Many of CGI Federal’s clients use cloud migration as the impetus to evolve the disaster recovery strategy with measurable results. For example, the move of the ERP system to the cloud enabled one CGI client to improve their recovery time objective by 96%. 

Taking advantage of the move to the cloud, a number of agencies have cost-effectively moved from a “bare bones” cold disaster recovery paradigm requiring restore of most of the environment from backups to a warm disaster recovery solution, with servers and storage ready to scale at the point of need. The benefit: recovery times of minutes or a few hours rather than several hours or days.  Agencies that require high availability or shorter RTOs may consider active-active setup with full redundancy of environments across Production and DR regions to support minimal to zero downtime. CGI’s cloud architects can readily assess the capabilities within the target cloud architecture, make recommendations, and articulate the tradeoffs between elements such as cost and risk. 

Moving ERP to the cloud with purpose

Federal ERP systems are vital to agency operations, and you must plan and execute the migration with great care and consideration. Agencies benefit from cloud architects who intimately understand federal ERP system architectures and cloud target reference architectures.

As cloud offerings continuously expand, architects must stay abreast of new services and on top of updates to reference architectures to offer agencies the best recommendations for cloud-based services. With expertise in federal ERP and federal cloud, CGI Federal’s architects stand ready to support your agency in optimizing the results of moving your ERP system to the cloud. Contact us to discuss your unique challenges and learn more about our approach to ERP cloud migration.