作者: Maryam Amiri , Leyli Mohammad-Khanli
DOI: 10.1016/J.JNCA.2017.01.016
关键词:
摘要: According to the dynamic nature of cloud and rapid growth resources demand in it, resource provisioning is one challenging problems environment. The should be allocated dynamically according changes application. Over-provisioning increases energy wasting costs. On other hand, under-provisioning causes Service Level Agreements (SLA) violation Quality (QoS) dropping. Therefore close current applications as much possible. Furthermore, speed response workload achieve desired performance level a critical issue for elasticity. For this purpose, future determined. Thus, prediction application different aspects (workload, performance) an essential step before provisioning. results, sufficient are appropriate time way that QoS ensured SLA avoided. This paper reviews state art methods aspects. Through meticulous literature review schemes, taxonomy models presented investigates main characteristics challenges models. Finally, open research issues trends discussed. HighlightsThe aspects.The presents models.It discusses prediction.