Performability evaluation: where it is and what lies ahead

Meyer, John F.

doi:10.1109/ipds.1995.395818

Cited by 32 publications

(17 citation statements)

References 75 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This extends the until path operator Φ U I J Ψ with an additional interval J which represents a bound for the accumulated rewards. Besides performability [20] and other important dependability measures, Cloth et. al.…”

Section: Handling Rewardsmentioning

confidence: 99%

Time-bounded model checking of infinite-state continuous-time Markov chains

Zhang

Hermanns

Hahn

et al. 2008

2008 8th International Conference on Application of Concurrency to System Design

View full text Add to dashboard Cite

show abstract

Section: Handling Rewardsmentioning

confidence: 99%

Time-bounded model checking of infinite-state continuous-time Markov chains

Zhang

Hermanns

Hahn

et al. 2008

2008 8th International Conference on Application of Concurrency to System Design

View full text Add to dashboard Cite

show abstract

“…The previous work on performability analysis [11], [22], [33] involved a rich set of stochastic process models that describe system dependencies, fault likelihoods over time, and performance. For example, Smith et al [33] construct a Markovian reward model representing the evolution of a multiprocessor system through states with different sets of operational components.…”

Section: Related Work 41 Availability and Performabilitymentioning

confidence: 99%

Quantifying the performability of cluster-based services

Nagaraja

Gama

Bianchini

et al. 2005

IEEE Trans. Parallel Distrib. Syst.

View full text Add to dashboard Cite

Abstract-In this paper, we propose a two-phase methodology for systematically evaluating the performability (performance and availability) of cluster-based Internet services. In the first phase, evaluators use a fault-injection infrastructure to characterize the service's behavior in the presence of faults. In the second phase, evaluators use an analytical model to combine an expected fault load with measurements from the first phase to assess the service's performability. Using this model, evaluators can study the service's sensitivity to different design decisions, fault rates, and other environmental factors. To demonstrate our methodology, we study the performability of a multitier Internet service. In particular, we evaluate the performance and availability of three soft state maintenance strategies for an online bookstore service in the presence of seven classes of faults. Among other interesting results, we clearly isolate the effect of different faults, showing that the tier of Web servers is responsible for an often dominant fraction of the service unavailability. Our results also demonstrate that storing the soft state in a database achieves better performability than storing it in main memory (even when the state is efficiently replicated) when we weight performance and availability equally. Based on our results, we conclude that service designers may want an unbalanced system in which they heavily load highly available components and leave more spare capacity for components that are likely to fail more often.

show abstract

“…Reliability and performance are often analyzed separately, but then the results either fail to capture aspects of performance, or depict performance in an ideal state. The notion of combining aspects of both reliability and performance is embodied in the concept of performability analysis ( [Mey95], [Col99]). Given that detailed reliability or performance analyses are difficult problems in their own right, performability analysis of necessity involves some simplification of both the reliability and performance aspects in order to be feasible.…”

Section: Introductionmentioning

confidence: 99%

Techniques for Probabilistic Multi-Layer Network Analysis

Oikonomou

Sinha

2008

IEEE GLOBECOM 2008 - 2008 IEEE Global Telecommunications Conference

View full text Add to dashboard Cite

We describe a prototype tool that analyzes combined aspects of reliability and performance for a variety of networks. Our network-wide analysis systematically generates failure scenarios (called network states), maps their effects from the physical to the network to the traffic layer, calculates the probability of each state, and evaluates a metric on it, until the expected value of the metric or a point on its distribution is estimated to sufficient accuracy.We describe an application of this multi-layer probabilistic methodology to dealing with partial link failures in an ISP backbone network. I. INTRODUCTIONAs communications networks grow in size and complexity, the evaluation of their reliability and performance becomes more and more critical. Network service providers usually guarantee levels of service in terms of down time, restoration delay, latency, etc., to their enterprise customers. These service-level agreements (SLAs), are stated as, for example, "with 99.99% probability, at most 5% of the network traffic will be unavailable".To determine whether a network meets such guarantees, it is important to have a methodology that can model and evaluate the impact of failures on both reliability and performance. Reliability and performance are often analyzed separately, but then the results either fail to capture aspects of performance, or depict performance in an ideal state. The notion of combining aspects of both reliability and performance is embodied in the concept of performability analysis ([Mey95], [Col99]). Given that detailed reliability or performance analyses are difficult problems in their own right, performability analysis of necessity involves some simplification of both the reliability and performance aspects in order to be feasible.Our view is that network performability analysis must involve the following essential ingredients:• Network-wide analysis, instead of e.g., studying the behavior of "reference" connections. • Hierarchical, multi-level network models, to capture the way failures really propagate through network layers. • Associating probabilities with failures, and with performance guarantees, instead of "deterministic" analyses. • Systematic exploration of the space of all possible failures, as opposed to only single failures, or what-if studies, for example. For this purpose we have developed nperf, a network performability analyzer. nperf examines both the reliability

show abstract

Performability evaluation: where it is and what lies ahead

Cited by 32 publications

References 75 publications

Time-bounded model checking of infinite-state continuous-time Markov chains

Time-bounded model checking of infinite-state continuous-time Markov chains

Quantifying the performability of cluster-based services

Techniques for Probabilistic Multi-Layer Network Analysis

Contact Info

Product

Resources

About