Dr Leandro Navarro
Universitat Politècnica de Catalunya
Department of Computer Architecture
Jordi Girona, 1-3, D6
ES - 08034 Barcelona
Email: Applicants are not required to contact the coordinator but just prepare and submit an application following the instructions in this web site.

Vamis Xhagjika

BarcelonaTech UPC, Spain

Kungliga Tekniska Högskolan KTH, Sweden

Joining Date: March 2013



Research Interests

cloud;computing;federation;structured federation;multi-cloud;self-healing;self-management;self-*;telecom infrastructure clouds;

Selected Publications

[1] V. Xhagjika, V. Vlassov, M. Molin, and S. Toma. Structured cloud federation for carrier and isp infrastructure. In Cloud Networking (CloudNet), 2014 IEEE 3rd International Conference on, pages 20–26, Oct 2014.

Cloud Computing in recent years has seen enhanced growth and extensive support by the research community and industry. The advent of cloud computing realized the concept of commodity computing, in which infrastructure (resources) can be allocated on demand giving the illusion of infinite resource availability. The state-of-art Carrier and ISP infrastructure technology is composed of tightly coupled software services with the underlying customized hardware architecture. The fast growth of cloud computing as a vastly consolidated and stabilized technology is appealing to Carrier Providers in order to reduce Carrier deployment costs and enable a future of Carrier Clouds with easily accessible virtual carriers. For such migration to happen software services need to be generalized, to decouple hardware and software, and prepared to move into the Cloud. The network backbone is centrally managed and only provides network connectivity. We believe this presents an opportunity. The edges of such networks and the core are interconnected with high performance links. If services could be deployed in these edges they would benefit from enhanced locality to the user. In this position paper we propose a distributed cloud architecture (precisely a structured multi-cloud federated infrastructure), with minimal impact on existing infrastructure, as a first step to incorporate the Cloud into the network infrastructure of such providers, enabling and enhancing novel and existing applications.

Keywords: cloud computing;resource allocation;ISP infrastructure;carrier clouds;carrier deployment cost reduction;carrier infrastructure;cloud computing;commodity computing;customized hardware architecture;distributed cloud architecture;resource allocation;software services;structured cloud federation;structured multicloud federated infrastructure;virtual carriers;Bandwidth;Cloud computing;Computer architecture;Computer integrated manufacturing;Network topology;Topology;Carrier;Cloud Architecture;Cloud Computing;Cloud Federation;IAAS;ISP;Network Infrastructure

[2] Ying Liu, V. Xhagjika, V. Vlassov, and A. Al Shishtawy. Bwman: Bandwidth manager for elastic services in the cloud. In Parallel and Distributed Processing with Applications (ISPA), 2014 IEEE International Symposium on, pages 217–224, Aug 2014.

The flexibility of Cloud computing allows elastic services to adapt to changes in workload patterns in order to achieve desired Service Level Objectives (SLOs) at a reduced cost. Typically, the service adapts to changes in workload by adding or removing service instances (VMs), which for stateful services will require moving data among instances. The SLOs of a distributed Cloud-based service are sensitive to the available network bandwidth, which is usually shared by multiple activities in a single service without being explicitly allocated and managed as a resource. We present the design and evaluation of BwMan, a network bandwidth manager for elastic services in the Cloud. BwMan predicts and performs the bandwidth allocation and tradeoffs between multiple service activities in order to meet service specific SLOs and policies. To make management decisions, BwMan uses statistical machine learning (SML) to build predictive models. This allows BwMan to arbitrate and allocate bandwidth dynamically among different activities to satisfy specified SLOs. We have implemented and evaluated BwMan for the OpenStack Swift store. Our evaluation shows the feasibility and effectiveness of our approach to bandwidth management in an elastic service. The experiments show that network bandwidth management by BwMan can reduce SLO violations in Swift by a factor of two or more.

Keywords: bandwidth allocation;cloud computing;learning (artificial intelligence);public domain software;BwMan;OpenStack Swift store;SLOs;SML;cloud computing;elastic services;network bandwidth management;service instances;service level objectives;statistical machine learning;Bandwidth;Channel allocation;Data models;Monitoring;Predictive models;Servers;Throughput;Bandwidth Management;Cloud Computing;SLO