Schedule processes, not VCPUs

Song, Xiang; Shi, Jicheng; Chen, Haibo; Zang, Binyu

doi:10.1145/2500727.2500736

Cited by 36 publications

(11 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, even this might be problematic in the context of virtualization: scheduling multiple virtual machines in the presence of locks is a well-known challenge [19,36,55,59]. Essentially, the guest OS does not have control over the scheduling decisions of the hypervisor.…”

Section: Unresponsive Threadsmentioning

confidence: 99%

Concurrent Search Data Structures Can Be Blocking and Practically Wait-Free

David

Guerraoui

2016

Proceedings of the 28th ACM Symposium on Parallelism in Algorithms and Architectures

View full text Add to dashboard Cite

We argue that there is virtually no practical situation in which one should seek a "theoretically wait-free" algorithm at the expense of a state-of-the-art blocking algorithm in the case of search data structures: blocking algorithms are simple, fast, and can be made "practically wait-free".We draw this conclusion based on the most exhaustive study of blocking search data structures to date. We consider (a) different search data structures of different sizes, (b) numerous uniform and non-uniform workloads, representative of a wide range of practical scenarios, with different percentages of update operations, (c) with and without delayed threads, (d) on different hardware technologies, including processors providing HTM instructions.We explain our claim that blocking search data structures are practically wait-free through an analogy with the birthday paradox, revealing that, in state-of-the-art algorithms implementing such data structures, the probability of conflicts is extremely small. When conflicts occur as a result of context switches and interrupts, we show that HTM-based locks enable blocking algorithms to cope with them.

show abstract

Section: Unresponsive Threadsmentioning

confidence: 99%

Concurrent Search Data Structures Can Be Blocking and Practically Wait-Free

David

Guerraoui

2016

Proceedings of the 28th ACM Symposium on Parallelism in Algorithms and Architectures

View full text Add to dashboard Cite

show abstract

“…This problem can be solved through prioritization of CPU resources by employing one of numerous mechanisms and algorithms that allow individual virtual servers to gain higher priority over the physical server's CPU resources. However, by establishing process execution priority, the number of processor cores still does not increase (Song et al, 2013;Guan et al, 2014). During our study of the literature, several problems related to the existing research were identified as follows:…”

Section: Existing Solutions and Related Workmentioning

confidence: 99%

Model for Automated and Improved Utilization of Existing Computer Resources on an Example of Web Servers

Alagić¹,

Magdalenić²

2018

Journal of Computer Science

View full text Add to dashboard Cite

Information technology is under constant innovation pressure to provide the highest level of data availability-i.e., the continuous functioning of operating systems. This is the very reason for an accelerated development of complex systems encompassed by the term cloud computing. Among other things, such solutions are aimed to ensure high-level availability of complex systems and architecture. Numerous studies indicate that the main cause of high cost of data centers are procurement and maintenance costs of servers. Furthermore, this paper presents concrete practice examples which confirm that there are systems whose resources are not sufficiently exploited, but still have to be provided due to their importance. The high costs and inadequate utilization of the existing computer resources constitute the main motivation for the present research. In this study, we present a new model for automated and improved utilization of the existing computing resources. The model is verified by using an application for automated management of computer resources that was developed for this research and by several tests conducted on Web clusters.

show abstract

“…This will significantly increase synchronization latency and block the progress of other VCPUs waiting to acquire the same lock. More details can be found in existing literatures [4,5,13].…”

Section: Design Motivationmentioning

confidence: 99%

“…Four functions are used to get a suitable VCPU when runnable VCPUs include parallel ones (lines [8][9][10][11][12][13][14][15][16][17][18][19]. First, it calls get_parallel_vcpus_with_maxPackets() to obtain VCPUs, which are put in vcpuSet1, from run queue runq if parallel VMs they belongs to receive maximum packets since their last de-scheduled (line 8).…”

Section: Svs Algorithmmentioning

confidence: 99%

Synchronization-Aware Scheduling for Virtual Clusters in Cloud

Chen

et al. 2015

IEEE Trans. Parallel Distrib. Syst.

View full text Add to dashboard Cite

Due to high flexibility and cost-effectiveness, cloud computing is increasingly being explored as an alternative to local clusters by academic and commercial users. Recent research already confirmed the feasibility of running tightly-coupled parallel applications with virtual clusters. However, such types of applications suffer from significant performance degradation, especially as the overcommitment is common in cloud. That is, the number of executable Virtual CPUs (VCPUs) is often larger than that of available Physical CPUs (PCPUs) in the system. The performance degradation is mainly due to the fact that the current Virtual Machine Monitors (VMMs) are unaware of the synchronization requirements of the VMs which are running parallel applications. In this paper, There are two key contributions. (1) We propose an autonomous synchronization-aware VM scheduling (SVS) algorithm, which can effectively mitigate the performance degradation of tightly-coupled parallel applications running atop them in overcommitted situation.(2) We integrate the SVS algorithm into Xen VMM scheduler, and rigorously implement a prototype. We evaluate our design on a real cluster environment with NPB benchmark and real-world trace. Experiments show that our solution attains better performance for tightly-coupled parallel applications than the state-of-the-art approaches like Xen's Credit scheduler, balance scheduling, and hybrid scheduling.

show abstract

Schedule processes, not VCPUs

Cited by 36 publications

References 10 publications

Concurrent Search Data Structures Can Be Blocking and Practically Wait-Free

Concurrent Search Data Structures Can Be Blocking and Practically Wait-Free

Model for Automated and Improved Utilization of Existing Computer Resources on an Example of Web Servers

Synchronization-Aware Scheduling for Virtual Clusters in Cloud

Contact Info

Product

Resources

About