| Titre : | A Fault Tolerance Approach based on Checkpointing in the Cloud Computing Environments |
| Auteurs : | Nadjet Kerbouche, Auteur ; Fatma Zohra Bouchra Hocini, Auteur ; Samir Setaouti, Directeur de thèse |
| Type de document : | texte manuscrit |
| Editeur : | Université mustapha stambouli de Mascara:Faculté des sciences exactes, 2022 |
| ISBN/ISSN/EAN : | SE02214T |
| Format : | 52P. / couv. ill. / 29cm. |
| Accompagnement : | disque optique numérique (CD-ROM) |
| Langues: | Anglais |
| Résumé : |
In recent years, Cloud Computing has been progressively adopted as the principal solution to meet the ever-increasing demand for computing, storage, and networking resources. It allows users to simply deploy their applications and control running expenses while delegating quality of service enforcement to the provider. This one, for its part, have to ensure that, in the case of a failure, no data is lost and that deployed apps continue to work flawlessly. This characteristic is critical because failures are common on the Cloud due to the large number of workstations and the frequent usage of existing hardware. In this context, many research efforts have been devoted to address the issue of failures and to make Clouds more reliable. Therefore, research on fault tolerance in Cloud Computing mainly focused on adapting and extending existing techniques, while taking into account Cloud specificities. Among these techniques is Checkpointing, often known as restart, it allows applications to continue running even when faults occur. This is accomplished by taking periodic checkpoints of the operating application state, which are preserved on secondary hardware (new virtual machines or external storage memories). As a result, if the primary system on which the application is executing fails, the application's state is not lost, and the application can be restored and executed again on a secondary machine from the latest stored state. In our work, we suggest an adaptive checkpointing interval approach to achieve high reliability of cloud computing services and to avoid SLA violation penalties that are arise when the customer's expected deadline is exceeded. This proposed approach contains two phases: the first phase is dedicated to select the appropriate virtual machines (VMs) for each task by calculating the expected response time, and taking into account penalties cost. In the second phase we focus on applying flexible length of checkpoint interval depending on response time, failure percentage of VMs, penalty cost and fidelity. To evaluate our proposed strategy, we developed a platform to launch a set of experiments by simulating various failure scenarios. Performance is measured using response time, penalty cost, failure percentage criteria. In the future work, we will try to include more parameters as storage, network bandwidth …etc we will try also to use an existing benchmark load, in the experimental setup. |
Exemplaires (1)
| Code-barres | Cote | Support | Localisation | Section | Disponibilité |
|---|---|---|---|---|---|
| SE02214T | INF701 | Livre audio | Bibliothèque des Sciences Exactes | 7-Mémoires Master | Consultation sur place Exclu du prêt |
Aucun avis, veuillez vous identifier pour ajouter le vôtre !
Accueil
