東華大學圖書館 |

Benchmarking Microservice Platforms and Applications in the Cloud = = Benchmarking von Microservice-Plattformen und Anwendungen in der Cloud de /

Record Type:	Electronic resources : Monograph/item
Title/Author:	Benchmarking Microservice Platforms and Applications in the Cloud =/ Martin Grambow.
Reminder of title:	Benchmarking von Microservice-Plattformen und Anwendungen in der Cloud de /
Author:	Grambow, Martin,
Description:	1 electronic resource (149 pages)
Notes:	Source: Dissertations Abstracts International, Volume: 86-04, Section: A.
Contained By:	Dissertations Abstracts International86-04A.
Subject:	Software. -
Online resource:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31609054
ISBN:	9798342132299

Benchmarking Microservice Platforms and Applications in the Cloud = = Benchmarking von Microservice-Plattformen und Anwendungen in der Cloud de /
Grambow, Martin,

Benchmarking Microservice Platforms and Applications in the Cloud =Benchmarking von Microservice-Plattformen und Anwendungen in der Cloud de /Martin Grambow. - 1 electronic resource (149 pages)

Source: Dissertations Abstracts International, Volume: 86-04, Section: A.

Benchmarking can be used for evaluating software performance metrics within a controlled environment, for example, to ensure the compliance with Service Level Agreements (SLAs). Benchmarking microservice platforms and applications in cloud environments, however, is challenging. First, there are performance variations in cloud environments that make it difficult to accurately determine performance metrics and replicate results. Consequently, there is a trade-off between accuracy on the one hand and effort on the other. Running longer and more frequent benchmarking experiments improves the accuracy when quantifying performance metrics and detecting performance changes. Second, running extensive benchmarking experiments after every code change is unpractical due to the continuous development of microservice applications. Extensive benchmarking experiments are time-consuming, complex, costly and slow down the rapid development cycles of microservice applications. Third, cloud platforms undergo constant updates as well, including new hardware or software implementations that directly impact the performance of microservice applications.In this thesis, we propose three approaches to simplify benchmarking of microservice platforms and applications in cloud environments. First, we propose an approach to automatically adapt and generate the benchmark workload for a microservice application based on abstract interaction patterns and machine-generated service description files. By applying this concept, benchmark workloads no longer need to be manually adjusted after each change in a microservice, but are automatically adapted. Second, we optimize microbenchmark suites based on function calls during an application benchmark and study their performance change detection capabilities in realistic Continuous Integration/Deployment (CI/CD) pipelines. This opens up a variety of application scenarios for CI/CD pipelines where, e.g., the optimized microbenchmark suite might scan the application for performance issues after every source code change while the more complex application benchmark is run only for major releases. Third, we present BeFaaS, an extensible application-centric benchmarking framework for Function as a Service (FaaS) platforms. We show that BeFaaS can be used to study typical use cases and analyze various performance parameters of FaaS platforms. Although the presented benchmarking approaches are primarily applicable to microservice applications and platforms in the cloud, they can also be used in other domains.

English

ISBN: 9798342132299Subjects--Topical Terms:

619355
Software.

Benchmarking Microservice Platforms and Applications in the Cloud = = Benchmarking von Microservice-Plattformen und Anwendungen in der Cloud de /
LDR:06944nmm a22004333i 4500 001 2431786
005 20250624130450.5
006 m o d
007 cr|nu||||||||
008 261011s2024 miu||||||m |||||||eng d
020 $a 9798342132299
035 $a (MiAaPQD)AAI31609054
035 $a (MiAaPQD)TechBerlin_20241130322186
035 $a AAI31609054
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Grambow, Martin, $e author. $3 3820754
245 1 0 $a Benchmarking Microservice Platforms and Applications in the Cloud = $b Benchmarking von Microservice-Plattformen und Anwendungen in der Cloud de / $c Martin Grambow.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2024
300 $a 1 electronic resource (149 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 86-04, Section: A.
500 $a Advisors: Bermbach, David; Sprekeler, Henning; Hasselbring, Wilhelm; Kao, Odej.
502 $b Ph.D. $c Technische Universitaet Berlin (Germany) $d 2024.
520 $a Benchmarking can be used for evaluating software performance metrics within a controlled environment, for example, to ensure the compliance with Service Level Agreements (SLAs). Benchmarking microservice platforms and applications in cloud environments, however, is challenging. First, there are performance variations in cloud environments that make it difficult to accurately determine performance metrics and replicate results. Consequently, there is a trade-off between accuracy on the one hand and effort on the other. Running longer and more frequent benchmarking experiments improves the accuracy when quantifying performance metrics and detecting performance changes. Second, running extensive benchmarking experiments after every code change is unpractical due to the continuous development of microservice applications. Extensive benchmarking experiments are time-consuming, complex, costly and slow down the rapid development cycles of microservice applications. Third, cloud platforms undergo constant updates as well, including new hardware or software implementations that directly impact the performance of microservice applications.In this thesis, we propose three approaches to simplify benchmarking of microservice platforms and applications in cloud environments. First, we propose an approach to automatically adapt and generate the benchmark workload for a microservice application based on abstract interaction patterns and machine-generated service description files. By applying this concept, benchmark workloads no longer need to be manually adjusted after each change in a microservice, but are automatically adapted. Second, we optimize microbenchmark suites based on function calls during an application benchmark and study their performance change detection capabilities in realistic Continuous Integration/Deployment (CI/CD) pipelines. This opens up a variety of application scenarios for CI/CD pipelines where, e.g., the optimized microbenchmark suite might scan the application for performance issues after every source code change while the more complex application benchmark is run only for major releases. Third, we present BeFaaS, an extensible application-centric benchmarking framework for Function as a Service (FaaS) platforms. We show that BeFaaS can be used to study typical use cases and analyze various performance parameters of FaaS platforms. Although the presented benchmarking approaches are primarily applicable to microservice applications and platforms in the cloud, they can also be used in other domains.
520 $a Benchmarking kann zur Bestimmung von Software-Leistungsmetriken in einer kontrollierten Umgebung verwendet werden, um beispielsweise die Einhaltung von SLAs zu gewahrleisten. Das Benchmarking von Microservice-Plattformen und -Anwendungen in Cloud-Umgebungen ist jedoch herausfordernd. Erstens gibt es in Cloud-Umgebungen Leistungsschwankungen, die eine genaue Bestimmung von Leistungsmetriken und die Replizierung von Ergebnissen erschweren. Folglich gibt es einen Kompromiss zwischen der Genauigkeit auf der einen Seite und dem Aufwand auf der anderen Seite. Die Durchfuhrung langerer und mehrmals wiederholter Benchmarking-Experimente verbessert die Genauigkeit bei der Bestimmung von Leistungsmetriken und der Erkennung von Leistungsanderungen. Zweitens ist die Durchfuhrung umfangreicher Benchmarking-Experimente nach jeder Codeanderung aufgrund der standigen Weiterentwicklung von Microservice-Anwendungen nicht praktikabel. Umfangreiche Benchmarking-Experimente sind zeitaufwandig, komplex, und kostspielig, was die schnellen Entwicklungszyklen von Microservice-Anwendungen bremst. Drittens werden auch die Cloud-Plattformen standig weiterentwickelt, was unter anderem auch Hardware- oder Softwareaktualisierungen beinhaltet, die sich direkt auf die Leistungen von Microservice-Anwendungen auswirken.In dieser Arbeit prasentieren wir drei Ansatze, die das Benchmarking von Microservice-Plattformen und -Anwendungen in der Cloud vereinfachen. Erstens, einen Ansatz der die Benchmark-Last fur eine Microservice-Anwendung automatisch basierend auf maschinengenerierten Dienstbeschreibungen und abstrakten Interaktionsmustern generiert. Durch die Anwendung dieses Konzepts muss die Last fur einen Benchmark nicht mehr manuell nach jeder Anderung in einem Microservice angepasst werden, sondern wird automatisch angepasst. Zweitens optimieren wir Microbenchmark-Suites auf der Grundlage von Funktionsaufrufen wahrend eines Anwendungsbenchmarks und untersuchen deren Erkennungsfahigkeiten von Leistungsanderungen in realistischen CI/CD-Pipelines. Dies eroffnet eine Vielzahl von Anwendungsszenarien fur CI/CDPipelines, in denen beispielsweise die optimierte Suite die Anwendung auf Leistungsprobleme nach jeder Code-Anderung oder Commit durchsucht, wahrend komplexere Benchmarks nur fur grosere Entwicklungsschritte ausgefuhrt werden. Drittens stellen wir BeFaaS vor, ein erweiterbares und anwendungsorientiertes Benchmarking-Framework fur FaaS-Plattformen. Wir zeigen, dass BeFaaS verwendet werden kann, um typische Anwendungsfalle zu untersuchen und verschiedene Leistungsparameter von FaaS-Plattformen zu analysieren. Die vorgestellten Ansatze sind hauptsachlich fur das Benchmarking von Microservice-Anwendungen und -Plattformen in der Cloud anwendbar, konnen aber auch in anderen Bereichen angewendet werden.
546 $a English
590 $a School code: 1119
650 4 $a Software. $2 gtt. $3 619355
650 4 $a Software development. $3 3768860
650 4 $a Application programming interface. $3 3562904
650 4 $a Benchmarks. $3 3548256
650 4 $a Cloud computing. $3 1016782
650 4 $a Computer science. $3 523869
650 4 $a Statistics. $3 517247
690 $a 0984
690 $a 0454
690 $a 0463
710 2 $a Technische Universitaet Berlin (Germany). $e degree granting institution. $3 3819406
720 1 $a Bermbach, David $e degree supervisor.
720 1 $a Sprekeler, Henning $e degree supervisor.
720 1 $a Hasselbring, Wilhelm $e degree supervisor.
720 1 $a Kao, Odej $e degree supervisor.
773 0 $t Dissertations Abstracts International $g 86-04A.
790 $a 1119
791 $a Ph.D.
792 $a 2024
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31609054