MORPHOSYS: Efficient Colocation of QoS-Constrained Workloads in the Cloud

Abstract

Abstract-In hosting environments such as IaaS clouds, desirable application performance is usually guaranteed through the use of Service Level Agreements (SLAs), which specify minimal fractions of resource capacities that must be allocated for unencumbered use for proper operation. Arbitrary colocation of applications with different SLAs on a single host may result in inefficient utilization of the host's resources. In this paper, we propose that periodic resource allocation and consumption models -often used to characterize real-time workloads -be used for a more granular expression of SLAs. Our proposed SLA model has the salient feature that it exposes flexibilities that enable the infrastructure provider to safely transform SLAs from one form to another for the purpose of achieving more efficient colocation. Towards that goal, we present MORPHOSYS: a framework for a service that allows the manipulation of SLAs to enable efficient colocation of arbitrary workloads in a dynamic setting. We present results from extensive trace-driven simulations of colocated Video-on-Demand servers in a cloud setting. These results show that potentially-significant reduction in wasted resources (by as much as 60%) are possible using MORPHOSYS.

Formal Verification of SLA Transformations

“…Along the first, we are employing AARTIFACT in proving additional conjectured theorems about other models of SLA transformations for workload colocation. We believe additional SLA transformations will enhance the efficiency of our colocation framework (e.g., in [11], SLA transformations such as those seen in this paper yield a reduction of up to 60% in wasted resources). Along the second, we are interested in using AARTIFACT to reason about other models of SLA "calculus" (e.g., an SLA transformation for cloud storage).…”

“…3 This necessitates the use of searching and pruning heuristics that explore the solution space -namely, what SLA transform(s) to apply to each task, and how to partition the resulting set of tasks. Details of heuristics that were shown to be fairly effective (achieving significant packing efficiencies when compared to random co-location) and practical (scaling up to dozens of tasks per host) are given in [5], [11].…”

Robust partial-load experiments with Showstopper

“…A large share of results on workload consolidation relies on simulation for evaluation [2,29,51,[53][54][55]60,62,63]. Coarse granularity simulation often uses processor utilization log replay to approximate realistic workloads, assuming additive character of processor utilization.…”