In large-scale distributed infrastructures such as clouds, Grids, peer-to-peer systems, and wide-area testbeds, users and administrators typically desire to perform "on-demand operations" that deal with the most up-to-date state of the infrastructure. However, the scale and dynamism present in the operating environment make it challenging to support on-demand operations efficiently, i.e., in a bandwidth- and response-efficient manner. This dissertation discusses several on-demand operations, challenges associated with them, and system designs that meet these challenges. Specifically, we design and implement techniques for (1) on-demand group monitoring that allows users and administrators of an infrastructure to query and aggregate the up-to-date state of the machines (e.g., CPU utilization) in one or multiple groups, (2) on-demand storage for intermediate data generated by dataflow programming paradigms running in clouds, (3) on-demand Grid scheduling that makes worker-centric scheduling decisions based on the current availability of compute nodes, and (4) on-demand key/value pair lookup that is overlay-independent and perturbation-resistant. We evaluate these on-demand operations using large-scale simulations with traces gathered from real systems, as well as via deployments over real testbeds such as Emulab and PlanetLab. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com.bibliotheek.ehb.be/en-US/products/dissertations/individuals.shtml.]