att_abstract={{A complex cloud application consists of virtual machines (VMs) running software such as web servers and load balancers, storage in the form of disk volumes, and network connections that enable communication between VMs and be- tween VMs and disk volumes. The application is also associated with various properties, including not just quantities such as the size of a VM or a disk volume, but also quality of service (QoS) attributes such as throughput, latency, and reliability. This paper presents Ostro, an OpenStack-based scheduler that optimizes the utilization of data center resources, while satisfying the properties associated with cloud applications. The novelty of the approach realized by Ostro is that it makes holistic placement decisions, in which all the requirements of an application are considered jointly using an application topology abstraction. Three specific placement algorithms for application topologies are described, in- cluding an estimate-based greedy algorithm and a time-bounded A* algorithm. These algorithms can deal with complex topologies that have heterogeneous resource requirements, while still being scalable enough to handle the placement of hundreds of VMs and volumes across several thousands of host servers. The approach is evaluated using both extensive simulations and realistic ex- periments. These results show that Ostro significantly improves resource utilization when compared with a naive approach.}},
	att_authors={gj6146, mh7921, kj2681, rp267p, rs2497},
	att_copyright_notice={{This version of the work is reprinted here with permission of IEEE for your personal use. Not for redistribution. The definitive version was published in 2015. {{, 2015-07-02}}
	att_tags={cloud,  optimization,  performance,  scalability},
	author={Gueyoung Jung and Matti Hiltunen and Kaustubh Joshi and Rajesh Panta and Richard Schlichting},
	institution={{IEEE International Conference on Distributed Computing Systems}},
	title={{Ostro: Scalable Placement Optimization of Complex Application Topologies in Large-Scale Data Centers}},