Credence Successfully Completes the Migration of Wide Area WorkFlow (WAWF) Suite of Applications to AWS GovCloud

March 2, 2021 — As a Managed Service Provider, Credence is providing the Defense Logistics Agency a range of advanced technical and program management under our enterprise Cloud Managed Services (CMS) support contract and migrated Wide Area WorkFlow (WAWF) suite of applications to AWS GovCloud. Credence has consistently delivered advanced technical and enterprise solutions to its government clients and looks forward to expanding its partnership with the Defense Logistics Agency on similar contracts of critical importance.

Problem Solution/Definition:
DLA hosted WAWF suite of applications in DISA managed data centers and had issues monitoring resource utilization, virtual machine provisioning, and meeting the needs of Agile software development teams including issues with application uptime and costs.

Proposed Solution & Architecture:
Credence team of Cloud Engineers and Architects migrated and now manage the suite of applications with absolutely no unplanned downtime of the applications. Credence team used AWS GovCloud infrastructure and developed metrics in collaboration with PIEE (WAWF) Program Management Office (PMO) and streamed data from AWS CloudWatch to train a generative adversarial neural network. This AI then powered the knowledge of the instance utilization behavior and is used to forecast us and recommend optimal provisioning.

Outcomes of Projects & Success Metrics:
Reduced Cost by 20% within 2 weeks after the launch without application downtime. Application functions were not affected due to migration. Query performance was improved by up to 30% as a result of migration to AWS GovCloud.

Describe TCO Analysis Performed:
AWS pricing calculator was used to determine cost of infrastructure used to host and run code.

Lessons Learned:
We learned the benefits of EC2 instance type optimization beyond base utilization scaling within a provisioned instance type. This led us to apply statistical models to predict and optimize provisioning of RDS as well.