Inferring Performance from Code: A Review

Incerto, Emilio; Napolitano, Annalisa; Tribastone, Mirco

doi:10.1007/978-3-030-61362-4_17

Cited by 3 publications

(2 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this case, Algorithm 2 can be used to generate positive tests for the new version using the old generator model. Otherwise, the discriminator model is trained with the newly executed tests and it will contest with the generator model until the FJD distance between the synthetic tests and the executed ones becomes low again (lines [13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28]. The contest procedure is the same as that of Algorithm 1.…”

Section: Acta Applied To Devopsmentioning

confidence: 99%

“…Performance is one of the fundamental software quality attributes, whose assessment is crucial to guarantee that a deployed software functions correctly [9], and it is perceived as the new correctness in many practical deployment scenarios [13]. Performance issues may lead to several problems from degradation of performance indicators (e.g.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Automated Performance Testing Based on Active Deep Learning

Sedaghatbaf¹,

Moghadam²,

Saadatmand³

2021

Preprint

View full text Add to dashboard Cite

Generating tests that can reveal performance issues in large and complex software systems within a reasonable amount of time is a challenging task. On one hand, there are numerous combinations of input data values to explore. On the other hand, we have a limited test budget to execute tests. What makes this task even more difficult is the lack of access to source code and the internal details of these systems. In this paper, we present an automated test generation method called ACTA for black-box performance testing. ACTA is based on active learning, which means that it does not require a large set of historical test data to learn about the performance characteristics of the system under test. Instead, it dynamically chooses the tests to execute using uncertainty sampling. ACTA relies on a conditional variant of generative adversarial networks, and facilitates specifying performance requirements in terms of conditions and generating tests that address those conditions. We have evaluated ACTA on a benchmark web application, and the experimental results indicate that this method is comparable with random testing, and two other machine learning methods, i.e. PerfXRL and DN.

show abstract