TY - JOUR
T1 - Networking for the Cloud
T2 - Challenges and Trends
AU - Drago, Idilio
AU - de Oliveira Schmidt, Ricardo
AU - Hofstede, Rick J.
AU - Sperotto, Anna
AU - Karimzadeh Motallebi Azar, Morteza
AU - Haverkort, Boudewijn R.H.M.
AU - Pras, Aiko
N1 - eemcs-eprint-24117
PY - 2013/12/11
Y1 - 2013/12/11
N2 - Cloud services have changed the way computing power is delivered to customers, by offering computing and storage capacity in remote data centers on demand over the Internet. The success of the cloud model, however, has not come without challenges. Cloud providers have repeatedly been related to reports of major failures, including outages and performance degradation. The internal network of cloud data centers has frequently been identified as a root-cause of these problems, showing that network provisioning and monitoring is still a major challenge for the deployment of cloud services. This paper argues that today’s technologies for measuring and monitoring Internet traffic could be applied in the context of the internal network of cloud data centers as well. To support that, we first show the suitability of flow-based traffic measurements for monitoring cloud services. Then, we present a case on bandwidth capacity provisioning to exemplify how flow-based measurements can be used to guarantee the performance of cloud services. Finally, we discuss future directions we believe will guide the development of new cloud services. We advocate that next generation cloud services will not only rely on the Internet as a means to reach users, but also influence how the Internet itself is organized. We illustrate this trend by describing our ongoing research on mobile clouds.
AB - Cloud services have changed the way computing power is delivered to customers, by offering computing and storage capacity in remote data centers on demand over the Internet. The success of the cloud model, however, has not come without challenges. Cloud providers have repeatedly been related to reports of major failures, including outages and performance degradation. The internal network of cloud data centers has frequently been identified as a root-cause of these problems, showing that network provisioning and monitoring is still a major challenge for the deployment of cloud services. This paper argues that today’s technologies for measuring and monitoring Internet traffic could be applied in the context of the internal network of cloud data centers as well. To support that, we first show the suitability of flow-based traffic measurements for monitoring cloud services. Then, we present a case on bandwidth capacity provisioning to exemplify how flow-based measurements can be used to guarantee the performance of cloud services. Finally, we discuss future directions we believe will guide the development of new cloud services. We advocate that next generation cloud services will not only rely on the Internet as a means to reach users, but also influence how the Internet itself is organized. We illustrate this trend by describing our ongoing research on mobile clouds.
KW - Cloud networks
KW - Data centers
KW - Cloud services
KW - Traffic measurements
KW - Link Dimensioning
U2 - 10.1515/pik-2013-0035
DO - 10.1515/pik-2013-0035
M3 - Article
SN - 0930-5157
VL - 36
SP - 207
EP - 214
JO - PIK: Praxis der Informationsverarbeitung und Kommunikation
JF - PIK: Praxis der Informationsverarbeitung und Kommunikation
IS - 4
ER -