Using Heat and Ceilometer to create an elastic OpenStack grid

Grid computing is a term for connecting computing resources together to solve large computational problems. Computational grids are used for a lot of computations within the high energy physics domain, where the amount of computing power required for some tasks is vastly more than a local computer can provide. This thesis investigates if cloud technology can be utilized to make an elastic computational grid, in order to get access to more resources that would otherwise be idle. Functional requirements were defined for creating a prototype capable of providing a virtualized environment that scales the amount of virtual machines up and down automatically based on the load on the system. A prototype was created to take advantage of the technology provided by cloud, and the prototype tested to see how it fulfills the functional requirements. Although one of the functional requirements was not achieved, the test results demonstrate that the technology has promising potential, but further work and testing needs to be done