Containerised Application Profiling and Classification Using Benchmarks
Along with the rise of cloud and edge computing has come a plethora of solutions that regard the deployment and operation of different types of applications in such environments. Infrastructure as a service (IaaS) providers offer a number of different hardware solutions to facilitate the needs of th...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-12-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/12/23/12374 |
| _version_ | 1827643390624268288 |
|---|---|
| author | Alexandros Psychas Phivos Dadamis Nikolaos Kapsoulis Antonios Litke Theodora Varvarigou |
| author_facet | Alexandros Psychas Phivos Dadamis Nikolaos Kapsoulis Antonios Litke Theodora Varvarigou |
| author_sort | Alexandros Psychas |
| collection | DOAJ |
| description | Along with the rise of cloud and edge computing has come a plethora of solutions that regard the deployment and operation of different types of applications in such environments. Infrastructure as a service (IaaS) providers offer a number of different hardware solutions to facilitate the needs of the growing number of distributed applications. It is critical in this landscape to be able to navigate and discover the best-suited infrastructure solution for the applications, taking into account not only the cost of operation but also the quality of service (QoS) required for any given application. The proposed solution has two main research developments: (a) the creation and optimisation of multidimensional vectors that represent the hardware usage profiles of an application, and (b) the assimilation of a machine learning classification algorithm, in order to create a system that can create hardware-agnostic profiles of a vast variety of containerised applications in terms of nature and computational needs and classify them to known benchmarks. Given that benchmarks are widely used to evaluate a system’s hardware capabilities, having a system that can help select which benchmarks best correlate to a given application can help an IaaS provider make a more informed decision or recommendation on the hardware solution, not in a broad sense, but based on the needs of a specific application. |
| first_indexed | 2024-03-09T17:52:44Z |
| format | Article |
| id | doaj.art-d685a6c313b44de69bcd6a10f456a093 |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-09T17:52:44Z |
| publishDate | 2022-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-d685a6c313b44de69bcd6a10f456a0932023-11-24T10:35:28ZengMDPI AGApplied Sciences2076-34172022-12-0112231237410.3390/app122312374Containerised Application Profiling and Classification Using BenchmarksAlexandros Psychas0Phivos Dadamis1Nikolaos Kapsoulis2Antonios Litke3Theodora Varvarigou4School of Electrical and Computer Engineering, National Technical University of Athens, 9 Ir. Politechniou Str., 157 72 Athens, GreeceSchool of Electrical and Computer Engineering, National Technical University of Athens, 9 Ir. Politechniou Str., 157 72 Athens, GreeceInnov-Acts Ltd., 6 Kolokotroni Str., Nicosia 1101, CyprusSchool of Electrical and Computer Engineering, National Technical University of Athens, 9 Ir. Politechniou Str., 157 72 Athens, GreeceSchool of Electrical and Computer Engineering, National Technical University of Athens, 9 Ir. Politechniou Str., 157 72 Athens, GreeceAlong with the rise of cloud and edge computing has come a plethora of solutions that regard the deployment and operation of different types of applications in such environments. Infrastructure as a service (IaaS) providers offer a number of different hardware solutions to facilitate the needs of the growing number of distributed applications. It is critical in this landscape to be able to navigate and discover the best-suited infrastructure solution for the applications, taking into account not only the cost of operation but also the quality of service (QoS) required for any given application. The proposed solution has two main research developments: (a) the creation and optimisation of multidimensional vectors that represent the hardware usage profiles of an application, and (b) the assimilation of a machine learning classification algorithm, in order to create a system that can create hardware-agnostic profiles of a vast variety of containerised applications in terms of nature and computational needs and classify them to known benchmarks. Given that benchmarks are widely used to evaluate a system’s hardware capabilities, having a system that can help select which benchmarks best correlate to a given application can help an IaaS provider make a more informed decision or recommendation on the hardware solution, not in a broad sense, but based on the needs of a specific application.https://www.mdpi.com/2076-3417/12/23/12374application profiling and classificationcontainerised applicationsmachine learning classification methodssupervised learninginfrastructure as a service management |
| spellingShingle | Alexandros Psychas Phivos Dadamis Nikolaos Kapsoulis Antonios Litke Theodora Varvarigou Containerised Application Profiling and Classification Using Benchmarks Applied Sciences application profiling and classification containerised applications machine learning classification methods supervised learning infrastructure as a service management |
| title | Containerised Application Profiling and Classification Using Benchmarks |
| title_full | Containerised Application Profiling and Classification Using Benchmarks |
| title_fullStr | Containerised Application Profiling and Classification Using Benchmarks |
| title_full_unstemmed | Containerised Application Profiling and Classification Using Benchmarks |
| title_short | Containerised Application Profiling and Classification Using Benchmarks |
| title_sort | containerised application profiling and classification using benchmarks |
| topic | application profiling and classification containerised applications machine learning classification methods supervised learning infrastructure as a service management |
| url | https://www.mdpi.com/2076-3417/12/23/12374 |
| work_keys_str_mv | AT alexandrospsychas containerisedapplicationprofilingandclassificationusingbenchmarks AT phivosdadamis containerisedapplicationprofilingandclassificationusingbenchmarks AT nikolaoskapsoulis containerisedapplicationprofilingandclassificationusingbenchmarks AT antonioslitke containerisedapplicationprofilingandclassificationusingbenchmarks AT theodoravarvarigou containerisedapplicationprofilingandclassificationusingbenchmarks |