Session details: Special Issue on Challenges in Software Performance
Notice bibliographique
Résumé
Software systems have become increasingly complex. Such complexity makes it extremely challenging to deliver systems that meet their performance requirements. In recent years, there have been several high profile instances of projects that failed due to performance-related issues, e.g., the rollout of the healthcare.gov Web site. This motivates the need for new approaches for software developers to address performance problems. This special issue presents 6 papers that address this theme. A brief summary of these papers follows. Gesvindr and Buhnova study the challenges of building Platform as a Service (PaaS) cloud applications that meet their performance requirements. Using two case studies involving the Microsoft Azure cloud platform, the authors show that the use of a cloud platform in itself does not guarantee good performance and scalability. The authors then identify several best practices for ensuring the performance and scalability of PaaS applications. The work of Heinrich considers the problem of managing\ the performance of software systems composed of thirdparty cloud services. The lack of control over such thirdparty services can cause deviations from performance objectives that can be hard to foresee and manage. The author outlines a model-based framework called iObserve that can help a system adapt and evolve so as to address such problems. Li, O'Brien, and Kihl address the lack of standard methodologies for experimental performance evaluation of software systems. The paper outlines a method for defining domainspecific methodologies for performance evaluation in a way that is consistent across domains. The authors demonstrate the utility of their approach using a cloud services case study. The paper by Grottke et al. considers the problem of handling cyber attacks on networked computer systems andconnected industrial control systems supporting critical infrastructuresuch as energy, water, and gas. Using analytic models, the authors study the performance of various attack mitigation approaches. In particular, the model shows how parameters such as malware scanning rates, network topology, and types of infections impact the effective processing capacity of a system. Rehmann et al. describe the performance monitoring in SAP HANA's continuous integration process. They describe how performance regression tests have been integrated in the delivery process and what services are available to detect and report performance anomalies. Being an extended version of a previous workshop paper, this paper now also presents the reporting of performance test results and anomaly detection. Finally, Nambiar et al. examine issues and experience with software performance, and consider the shortcomings and potential benefits of modeling from the perspective of a potential industrial user. In particular, it considers why modeling is little used, and what conditions it would have to satisfy to fulfill its potential. We sincerely thank all authors for their work and contribution to this special issue. Additionally, we thank our reviewers Davide Arcelli, Danilo Ardagna, Alberto Avritzer, Steffen Becker, Ian Gorton, Andre van Hoorn, Raffaela Mirandola, Juan F. Perez, and Murray Woodside for their timely and constructive reviews as well as editor-in-chief Nidhi Hedge for her support.
Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.
Comment cette classification a été obtenuedéplier
Prédiction machine sur la base complète
Imitation des enseignantsNi prévalence calibrée, ni vérité terrain. Validation humaine à venir. Le volet Gemma est une étiquette directe du modèle pour chaque travail de la base, lue sur la notice réduite au titre. Le volet Codex est un classifieur appris des 10 348 étiquettes directes de Codex et calibré sur les taux pondérés de l'échantillon; les champs sans appui suffisant ne portent aucun appel Codex. Le mode candidate est l'union des deux volets; le consensus est leur intersection. Ces sorties portent le statut machine_predicted_unvalidated et ne sont pas des étiquettes humaines.
Scores du classifieur distillé par catégorie (deux têtes)
| Catégorie | Codex | Gemma |
|---|---|---|
| Métarecherche | 0,004 | 0,008 |
| Méta-épidémiologie (sens strict) | 0,003 | 0,001 |
| Méta-épidémiologie (sens large) | 0,003 | 0,002 |
| Bibliométrie | 0,003 | 0,002 |
| Études des sciences et des technologies | 0,003 | 0,001 |
| Communication savante | 0,011 | 0,009 |
| Science ouverte | 0,004 | 0,005 |
| Intégrité de la recherche | 0,009 | 0,008 |
| Charge utile insuffisante (le modèle a refusé de juger) | 0,326 | 0,175 |
Scores machine (provisoires)
Les deux têtes enseignantes du modèle étudiant, lues sur ce travail. Un score ordonne la base pour la relecture; il n'affirme jamais une catégorie, et le statut de validation accompagne chaque rangée tel quel.
Scores de référence d'un modèle non mature (critères de maturité non atteints, 7 itérations). Un score ordonne; il n'affirme jamais une catégorie.
score_only:v0-immature-baseline · tel quel depuis la passe de notation : score_only signifie que le nombre peut ordonner les travaux, et qu'aucune étiquette de catégorie n'en découleClassification
machine, non validéePrédiction automatique; un appel candidat d’une seule source (Gemma direct ou Codex distillé), pas un consensus.
Le détail, modèle par modèle et score par score, se trouve en fin de page sous « Comment cette classification a été obtenue ».