Designed scalable cloud solutions using OpenStack, Kubernetes, Ceph and SDN. Each cloud deployed has had outstanding uptime, great performance, and excellent cost effectiveness.Supported those designs by building extremely scalable Layer 3 Clos fabrics on merchant silicon. These networks leverage Cumulus Linux, BGP Unnumbered and Ansible to minimize administration effort. Designed and built a beautiful alert-monitoring system with Prometheus and Grafana. This system kept Awnix support in a proactive stance, presented information with focus on problem resolution, and was a key factor in consistently delivering outstanding uptime.

Architected, designed, developed, and demonstrated a state of the art hosting environment capable of automated deployment of virtual data centers. The model developed, including network, storage, and compute components will allow the delivery of a suite of services exceeding the capabilities of the current largest private cloud provider at a drastically reduced price point.

Deployed and tested VMware Site Recovery Manager enabling one-click disaster recovery for the PaaS/IaaS VMware environment.Engaged other branches at NITC to build a virtual environment to perform load testing of Dept. of Labor web applications. Enabling the testing outside of the heavily loaded production environment lead to many fixes that otherwise wouldn't have been feasible to discover and fix. Metrics taken from the testing environment paved the way for new business offerings in IBM WebSphere, JBoss, and IBM WebSphere Portal on VMware.Designed an HP Blade infrastructure to minimize cost of deploying multiple new VMware-based initiatives including a Citrix VDI cluster, VMware vCloud Director, and PaaS/IaaS hardware refresh. The design had a very tight schedule to be considered for funding, so a small team of personnel with HP BladeCenter experience was quickly formed to work out a design. With the design completed, HP engineering was brought in and able to quickly assess and fine-tune the design. The complete design was turned in on schedule.Facilitated work on VMs supporting National Defense University’s Virtual Worlds initiative. Leveraging the rapid provisioning, snapshot, and cloning capabilities of VMware, NITC system integrators were able to quickly test and revert changes from early-lifecycle software, minimizing the effect of many trial-and-error prone deployments. Vendors commented on how quickly NITC was able to test and report on new software versions.Supported migration of USDA Foreign Agriculture Service (FAS), Economic Research Service (ERS), and Grain Inspection Packers and Stockyards Administration (GIPSA) providing infrastructure and assistance in moving their applications and workloads to the NITC Kansas City Enterprise Data Center.

Designed and implemented an enterprise class VMware ESXi 3.5 cluster in 1 week with 1 week lead time for planning, allowing at the USDA National Information Technology Center (NITC) to attract business hosting the USDA's web site for the American Recovery and Reinvestment Act. The ARRA web site was driven by a backend application spanning multiple platforms, requiring the new VMware cluster to have significant integration with existing systems and infrastructure. This high visibility initiative was executed with alacrity and precision, involving the formation of a multi-disciplinary team of subject matter experts across six branches of the organization. The team leveraged the success of the ARRA site implementation and the availability of remaining resources in the VMware cluster to attract new business with a Platform as a Service (PaaS) offering.Designed and constructed a 4th generation VMware cluster for NITC's expanding PaaS offering. This VMware cluster improves upon the previous PaaS offering with a fault-tolerant, hot-add, horizontally scalable network design, leveraging 802.3ad channel bonding and 802.1q VLAN trunking. The storage subsystem has matured to a 4-way fault-tolerant, load-balanced 16Gbps connection per host to NITC's storage area network. Virtual machine data protection has been upgraded to a snapshot-based, full-image backup system allowing for in-parallel virtual machine recovery and cut-over.Constructed a VMWare cluster based on the 4th generation PaaS design for use as a remote disaster recovery system to allow off-site hosting of mission-critical systems. This cluster was built utilizing the resources of multiple branches of NITC and in accordance with all the proper steps of design, testing, and implementation before being shipped to the remote site. On-site implementation was accomplished in two days, including travel time to the remote site.