| |
| |
PANELS
Panel 1(PA1) - Tuesday 13:30-15:30 - Room: JM Synge
Analytics and Prediction in Systems and Service Management
Moderator: Nikos Anerousis, Research Staff Member and Manager, IBM TJ Watson Research Center
Panelists:
-
Claudio Bartolini, HP
-
Olivier Verscheure, IBM
-
Jean-Francois Huard, Netuitive
-
Prof. Charles Sutton, University of Edinburgh
-
Prof. Luciano Paschoal Gaspary, Universidade Federal do Rio Grande do Sul
Abstract
As organizations collect increasing amounts of data from their operations, there is increasing interest in applying analytical techniques to mine and interpret this data: discover patterns, understand the causes of outages, and drive smarter management and operational decisions. Particularly promising is the field of predictive analytics, which seeks to develop models where present and historical data can be used to predict future outcomes. There are several applications of this technology in the broader field of systems and network management, including forecasting the future state of managed elements, quantifying the risks of failure and other adverse outcomes, and taking early mitigation actions.
This panel discussion will explore the state of the art in analytical techniques applied to systems and service management. The panelists will present selected use cases of truly innovative management solutions that rely in predictive analytics and the discussion that will follow will attempt to establish connections with current management practices, as well as outline the opportunities that lie ahead.
Topics for further discussion:
- Examples of successful applications of analytics in management
- What lies ahead
- How management tools and techniques need to be enhanced to enable analytics-driven decision support
- What are the challenges and the obstacles to success
- What we need to do more
Panel 2 (PA2) - Wednesday 13:30-15:30 - Room: JM Synge
IT/Telco Convergence and the Maturation of Internet-scale Enterprise Applications
Moderator: Philip Robinson SAP Research, Belfast, Ireland
Panelists:
-
Sergi Figuerola (i2CAT, Barcelona)
-
Pascale Vicat-Blanc (LYATISS, Lyon)
-
Philip Taylor (SAP Research, Belfast)
-
Jayne Brady (Intune Networks, Belfast, UK)
-
Nicloa Ciulli (Nextworks, Pisa, Italy)
-
Sam Samuel (Bell-Labs/Alcatel-Lucent, Ireland)
Abstract
An Internet-scale Enterprise Application consists of globally-distributed components and subsystems delivered as services. Examples include Analytics, Logistics, Supply-Chain Management and Collaborative business. These services function autonomously with globally-distributed users, who have different workloads and connection characteristics including mobility, and globally distributed data sources. The volumes of stored and transmitted information continue to grow to the order of Petabytes, making intelligent resource usage optimization and communications acceleration techniques mandatory, including caching, compression, parallelisation and load-balancing. These applications will continue to evolve in features, size and usage as the associated business requirements change. The availability, performance, security and cost-effectiveness of these applications nevertheless remain critical, as they support business decisions and data in a fast-paced, economy-driven environment. The title “Internet-scale” is misleading, as the sufficiency of the existing best-effort Internet as the infrastructure and engine that enables these applications is questionable. Current cloud software and infrastructure providers deliver their products over the Internet but are already losing potential business as Enterprises express concerns regarding the reliability and security of Internet traffic and interactions. What are the infrastructural, technological and business steps that need to be taken in order to mature the capability to optimally deliver and consume Internet-scale Enterprise Applications?
Target Audience:
The panel session is expected to bring together business and technical aspects of Enterprise and Telecom systems. The scope of the audience that will benefit from this session is hence rather broad. An appreciation of the problems of non-convergence between IT and Telco system management from either of these perspectives would be an asset. However, such an understanding, along with the realization of steps being taken to address these issues, should be the take-away from this session.
Panel 3 (PA3) - Thursday 11:00-12:30 - Room: JM Synge
Next Generation Services Overlay Networks (NGSON)
Moderator: Mehmet Ulema, Manhattan College, NY, USA.
Panelists:
-
Rick Townsend, Chair of IEEE Project 1903 NGSON, USA
-
Raouf Boutaba, University of Waterloo, Canada
-
Li-Yan, Huawei Technologies Co., P.R.China
-
Mehmet Ulema, Manhattan College, USA
Abstract
Next generation service overlay network (NGSON) provides a framework and necessary functions to support context-aware, dynamically adaptive, and self-organizing networks over Internet Protocol (IP)-based networks. NGSON is envisioned to include advanced service and transport level routing and forwarding schemes that are independent of the underlying networks such as IMS, NGN, P2P overlay or WEB to transmit NGSON signaling messages and/or media among its users and services. Recognizing he importance of this issue, many standards organizations have been addressing various aspects of the issue under different frameworks and terminology. NGSON is the IEEE’s official name for this effort under IEEE Project 1903. A number of equipment vendors, telecom operators, and research organizations have been working on this project to standardize various aspects under P 1903. As of this writing, the Working Group is about to wrap up the functional architecture for NGSON.
The panelists will discuss various aspects of NGSON, as well as the status of NGSON standardization effort. Panelists will be invited to address any of, but not limited to, the following or related topic areas:
• NGSON architecture for wireline and wireless networks
• Status of NGSON standards
• Status of, and experiences with, NGSON implementation and testbed
• NGSON applications and services for both wireline and wireless IP-based networks
• Security issues with NGSON
• NGSON Service Routing
• NGSON Service Composition
• Dynamic Adaptation in NGSON
• Context Awareness in NGSON
• Self-Organization in NGSON
• QoS and Performance Issues in NGSON
• NGSON Network and Service Management
• NGSON interactions with IMS, P2P, etc
Panel 4 (PA4) - PLENARY PANEL - Thursday 16:00-18:00 - Room: Burke
SLA ready Secure Cloud – How do we get there?
Moderator: Masum Hasan, Cisco, USA
Panelists:
-
Prof. Volker Sander, FH Aachen, University of Applied Sciences
-
Gerard Parr, University of Ulster, Northern Ireland
-
Dr. Marc Lacoste, Orange Lab
-
Robert Knaap, EMC Ireland
-
Victor Reijs, HEAnet, Ireland
-
Ehab Al-Shaer, UNC Charlotte
Abstract
The term SLA typically is associated with either availability or QoS. But in a Cloud, SLA should assume expanded dimensions for the following major reasons:
-
Multiple types of resources (compute, storage and network) are offered in a Cloud as an integrated whole, such as a Virtual DC. Hence SLA should incorporate metrics from these resources, not in isolation, but in integrated fashion.
-
End-to-end (E2E) SLA: from customer premise via the (IP-NGN) MAN/WAN to one or more Cloud DCs to VMs to Storage.
-
Security now becomes integral part of SLA, since the resources acquired in the Cloud are no more on premises (in the enterprise). Hence E2E Security SLA has to be supported at various levels, from compute/data security to policy/regulatory-based security to traffic / route isolation in the network.
-
An Enterprise (Private Cloud) may extend into a Public or Community Cloud thus forming a Hybrid Cloud. SLA requirements in this case assumes a new meaning.
-
As on-premises enterprise applications or whole enterprise DC (Virtual DC) are moved to the Cloud, SLA requirements may assume new dimensions.
-
Depending on what type of applications (such as healthcare/financial/high-performance computing/realtime/mission-critical) move to Cloud, SLA requirement will be dependent on that .
|
|
|
| |
|
 
|
