A computer-implemented method, carried out by one or more processors, for managing resources in a server environment. The method includes determining, by one or more processors, to shut down a first resource consumer, wherein the first resource consumer is assigned a first virtual resource with a first set of one or more host resources. It is determined, by one or more processors, whether a second virtual resource assigned to a second resource consumer requires the first set of one or more host resources. If the second virtual resource assigned to the second resource consumer does not require the first set of one or more host resources, it is determined, by one or more processors, not to deactivate the one or more host resources assigned to the first virtual resource.