Omega

Schwarzkopf, Malte; Konwinski, Andy; Abd-El-Malek, Michael; Wilkes, John

doi:10.1145/2465351.2465386

Cited by 486 publications

(43 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…These results show a low prediction error (within 5%) for five out of seven applications and indicate that updating the model parameters online with the Model Updater improves accuracy for all the workloads compared to offline estimates of the optimal static parameters. Our results compare favorably with previous work leveraging a similar resource-performance model [Sharifi et al 2011] (see Section 5).…”

Section: Prediction Error On Single-vm Workloadssupporting

confidence: 84%

“…We rely on the presence of an admission control system that chooses VM placement across the data center [Chen et al 2012;Schwarzkopf et al 2013], and we assume that the hypervisor supports live migration [Jo et al 2013] in case a node becomes overloaded. While both VM placement and migration present open questions for research, AutoPro focuses on a different challenging and interesting problem: automating allocation of a contended resource within a single node, based on application-level performance SLOs.…”

Section: Automated Fine-grained Provisioningmentioning

confidence: 99%

“…ControlWare [Zhang et al 2002], METE [Sharifi et al 2011], and PTRADE [Hoffmann et al 2013] served as inspiration for subsystems of AutoPro. These systems tackle environments that differ from virtualization infrastructures, making a full-system comparison with AutoPro unfeasible; instead, we outline quantitative comparisons between subsystems.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Automated Fine-Grained CPU Provisioning for Virtual Machines

Bartolini

Sironi

Sciuto

et al. 2014

ACM Trans. Archit. Code Optim.

View full text Add to dashboard Cite

Ideally, the pay-as-you-go model of Infrastructure as a Service (IaaS) clouds should enable users to rent just enough resources (e.g., CPU or memory bandwidth) to fulfill their service level objectives (SLOs). Achieving this goal is hard on current IaaS offers, which require users to explicitly specify the amount of resources to reserve; this requirement is nontrivial for users, because estimating the amount of resources needed to attain application-level SLOs is often complex, especially when resources are virtualized and the service provider colocates virtual machines (VMs) on host nodes. For this reason, users who deploy VMs subject to SLOs are usually prone to overprovisioning resources, thus resulting in inflated business costs.This article tackles this issue with AutoPro: a runtime system that enhances IaaS clouds with automated and fine-grained resource provisioning based on performance SLOs. Our main contribution with AutoPro is filling the gap between application-level performance SLOs and allocation of a contended resource, without requiring explicit reservations from users. In this article, we focus on CPU bandwidth allocation to throughput-driven, compute-intensive multithreaded applications colocated on a multicore processor; we show that a theoretically sound, yet simple, control strategy can enable automated fine-grained allocation of this contended resource, without the need for offline profiling. Additionally, AutoPro helps service providers optimize infrastructure utilization by provisioning idle resources to best-effort workloads, so as to maximize node-level utilization.Our extensive experimental evaluation confirms that AutoPro is able to automatically determine and enforce allocations to meet performance SLOs while maximizing node-level utilization by supporting batch workloads on a best-effort basis.

show abstract

Section: Prediction Error On Single-vm Workloadssupporting

confidence: 84%

Section: Automated Fine-grained Provisioningmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Automated Fine-Grained CPU Provisioning for Virtual Machines

Bartolini

Sironi

Sciuto

et al. 2014

ACM Trans. Archit. Code Optim.

View full text Add to dashboard Cite

show abstract

“…Since multiple services commonly share servers in both public and private datacenters [Dean and Barroso 2013;Hindman et al 2011;Schwarzkopf et al 2013], there is a need for isolation between applications. The use of kernel-based or hypervisor-based networking stacks largely addresses the problem.…”

Section: Challenges For Datacenter Applicationsmentioning

confidence: 99%

The IX Operating System

Belay¹,

Prekas

Primorac

et al. 2016

ACM Trans. Comput. Syst.

View full text Add to dashboard Cite

The conventional wisdom is that aggressive networking requirements, such as high packet rates for small messages and μs-scale tail latency, are best addressed outside the kernel, in a user-level networking stack. We present IX, a dataplane operating system that provides high I/O performance and high resource efficiency while maintaining the protection and isolation benefits of existing kernels.IX uses hardware virtualization to separate management and scheduling functions of the kernel (control plane) from network processing (dataplane). The dataplane architecture builds upon a native, zero-copy API and optimizes for both bandwidth and latency by dedicating hardware threads and networking queues to dataplane instances, processing bounded batches of packets to completion, and eliminating coherence traffic and multicore synchronization. The control plane dynamically adjusts core allocations and voltage/frequency settings to meet service-level objectives.We demonstrate that IX outperforms Linux and a user-space network stack significantly in both throughput and end-to-end latency. Moreover, IX improves the throughput of a widely deployed, key-value store by up to 6.4× and reduces tail latency by more than 2×. With three varying load patterns, the control plane saves 46%-54% of processor energy, and it allows background jobs to run at 35%-47% of their standalone throughput.

show abstract

“…Monolithic cloud orchestrators have a tendency to become very complex [3]. This results hard-to-maintain and hard-to-scale bottlenecks.…”

Section: Introductionmentioning

confidence: 99%

Optimizing the Integration of Agent-Based Cloud Orchestrators and Higher-Level Workloads

Sebrechts

Seghbroeck

Turck

2017

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. The flexibility of cloud computing has put significant strain on operations teams. Manually installing and configuring applications in the cloud simply isn't an option anymore. Configuration management automation solves the issue of getting a single application into a certain state automatically and reliably. However, the issue of automatic dependency management between multiple applications is still an "open, hard problem" according to researchers at Google. Agent-based modeling and orchestration tools like Juju solve the issue of getting from zero to a working set of correctly clustered and connected frameworks. The shortcomings of these state-of-the-art tools are that they don't provide efficient ways to model and orchestrate workloads running on top of these frameworks. This paper presents a number of ways to deploy and orchestrate workloads with Juju, compares their performance and overhead, and suggests how this overhead can be minimized.

show abstract

Omega

Cited by 486 publications

References 17 publications

Automated Fine-Grained CPU Provisioning for Virtual Machines

Automated Fine-Grained CPU Provisioning for Virtual Machines

The IX Operating System

Optimizing the Integration of Agent-Based Cloud Orchestrators and Higher-Level Workloads

Contact Info

Product

Resources

About