A method of modeling and analyzing load across servers in a cloud computing environment. The method comprises analyzing a processor, a memory, and an input/output utilization of each of a plurality of servers, where each server is comprised of at least one logical partition. A chart displaying a usage pattern is presented for each server. A list of logical partitions is presented for each server of the plurality of servers. A selection of at least one logical partition on a first server of the plurality of servers and a selection of a second server of the plurality of servers is received. The utilization impacts of moving the logical partition to the selected second server are modeled and an updated chart displaying the usage pattern for each server based on the modeled migration of the logical partition is presented. The modeled optimal location stored in a disaster recovery plan.