Inside-Out: Reliable Performance Prediction for Distributed Storage Systems in the Cloud

Hsu, Chin-Jung; Panta, Rajesh; Ra, Moo-Ryong; Freeh, Vincent W.

doi:10.1109/srds.2016.025

Cited by 8 publications

(15 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Low-level performance metrics are a good proxy for estimating application and system performance [5], [1]. They are also useful to identify performance anomalies [11], [12].…”

Section: Percentage Of Workloadsmentioning

confidence: 99%

“…Prior work has shown how low-level performance metrics of workload are information, which is a good proxy for predicting performance [12], [5], [17], [1]. For example, the memory commit size represents the amount of memory required to handle current workload, and the CPU time waiting on I/O indicates the workload type or a bottleneck on I/O.…”

Section: A Choosing the Low-level Metricsmentioning

confidence: 99%

See 1 more Smart Citation

Arrow: Low-Level Augmented Bayesian Optimization for Finding the Best Cloud VM

Hsu

Nair

Freeh

et al. 2018

2018 IEEE 38th International Conference on Distributed Computing Systems (ICDCS)

Self Cite

View full text Add to dashboard Cite

With the advent of big data applications, which tends to have longer execution time, choosing the right cloud VM to run these applications has significant performance as well as economic implications. For example, in our large-scale empirical study of 107 different workloads on three popular big data systems, we found that a wrong choice can lead to a 20 times slowdown or an increase in cost by 10 times.Bayesian optimization is a technique for optimizing expensive (black-box) functions. Previous attempts have only used instancelevel information (such as # of cores, memory size) which is not sufficient to represent the search space. In this work, we discover that this may lead to the fragility problem-either incurs high search cost or finds only the sub-optimal solution. The central insight of this paper is to use low-level performance information to augment the process of Bayesian Optimization. Our novel low-level augmented Bayesian Optimization is rarely worse than current practices and often performs much better (in 46 of 107 cases). Further, it significantly reduces the search cost in nearly half of our case studies.Based on this work, we conclude that it is often insufficient to use general-purpose off-the-shelf methods for configuring cloud instances without augmenting those methods with essential systems knowledge such as CPU utilization, working memory size and I/O wait time.

show abstract

“…Low-level performance metrics are a good proxy for estimating application and system performance [5], [1]. They are also useful to identify performance anomalies [11], [12].…”

Section: Percentage Of Workloadsmentioning

confidence: 99%

Section: A Choosing the Low-level Metricsmentioning

confidence: 99%

Arrow: Low-Level Augmented Bayesian Optimization for Finding the Best Cloud VM

Hsu

Nair

Freeh

et al. 2018

2018 IEEE 38th International Conference on Distributed Computing Systems (ICDCS)

Self Cite

View full text Add to dashboard Cite

show abstract

“…Choosing the right cloud configuration is essential to maximize application performance and minimize operational costs. However, such optimization task is not straightforward due to opaque resource requirement [1], [2]. To address this challenge, prior work either builds prediction models (as in Ernest [3] and PARIS [1]) or uses sequential model-based optimization (as in CherryPick [4], Arrow [5] and Scout [6]).…”

Section: Introductionmentioning

confidence: 99%

Micky: A Cheaper Alternative for Selecting Cloud Instances

Hsu

Nair

Menzies

et al. 2018

2018 IEEE 11th International Conference on Cloud Computing (CLOUD)

Self Cite

View full text Add to dashboard Cite

Most cloud computing optimizers explore and improve one workload at a time. When optimizing many workloads, the single-optimizer approach can be prohibitively expensive. Accordingly, we examine "collective optimizer" that concurrently explore and improve a set of workloads significantly reducing the measurement costs. Our large-scale empirical study shows that there is often a single cloud configuration which is surprisingly near-optimal for most workloads. Consequently, we create a collective-optimizer, MICKY, that reformulates the task of finding the near-optimal cloud configuration as a multi-armed bandit problem. MICKY efficiently balances exploration (of new cloud configurations) and exploitation (of known good cloud configuration). Our experiments show that MICKY can achieve on average 8.6 times reduction in measurement cost as compared to the state-of-the-art method while finding near-optimal solutions.Hence we propose MICKY as the basis of a practical collective optimization method for finding good cloud configurations (based on various constraints such as budget and tolerance to nearoptimal configurations).1 Micky (Rosa) is a character, from the Hollywood movie 21, who founded the MIT Black Jack team of card counters.

show abstract

“…The system monitor collects, aggregates, and transforms unstructured storage metrics and statistics into general and valuable monitoring data [30,36,115]. It captures relevant properties of the physical storage environment, including both static and dynamic metrics (e.g., maximum disk and network switch performance, available storage space, IOPS, bandwidth usage) from SDS controllers and data plane tiers, and collects samples of I/O workloads, in order to trace an up-to-date profile of the storage stack.…”

Section: Control Plane -Controllersmentioning

confidence: 99%

A Survey and Classification of Software-Defined Storage Systems

et al. 2020

View full text Add to dashboard Cite

The exponential growth of digital information is imposing increasing scale and efficiency demands on modern storage infrastructures. As infrastructure complexity increases, so does the difficulty in ensuring quality of service, maintainability, and resource fairness, raising unprecedented performance, scalability, and programmability challenges. Software-Defined Storage (SDS) addresses these challenges by cleanly disentangling control and data flows, easing management, and improving control functionality of conventional storage systems. Despite its momentum in the research community, many aspects of the paradigm are still unclear, undefined, and unexplored, leading to misunderstandings that hamper the research and development of novel SDS technologies. In this article, we present an in-depth study of SDS systems, providing a thorough description and categorization of each plane of functionality. Further, we propose a taxonomy and classification of existing SDS solutions according to different criteria. Finally, we provide key insights about the paradigm and discuss potential future research directions for the field.

show abstract

Inside-Out: Reliable Performance Prediction for Distributed Storage Systems in the Cloud

Cited by 8 publications

References 19 publications

Arrow: Low-Level Augmented Bayesian Optimization for Finding the Best Cloud VM

Arrow: Low-Level Augmented Bayesian Optimization for Finding the Best Cloud VM

Micky: A Cheaper Alternative for Selecting Cloud Instances

A Survey and Classification of Software-Defined Storage Systems

Contact Info

Product

Resources

About