Transcript title
17/07/2015 A necessary symbiosis; Cloud Computing, IoT, Big Data and Mobile Brussels, 17th April Josep Martrat Atos [email protected] Your business technologists. Powering progress 1 17/07/2015 Agenda ▶ Atos, the worldwide IT partner ▶ Trends and context ▶ Cloud and Big data: advantages & barriers ▶ Big data options: storage & processing ▶ And mobiles and IoT comes into arena ▶ Challenges ▶ Scenarios examples 2 Important economic and IT trends are shaping a “new transformation" ▶ Main IT trends ▶ Main economic trends Users & applications Mobility and Internet of Things 17/07/2015 Economic power changing towards emerging economies Social Networks and Media The debt crisis leads to cost pressure Big Data Cloud Computing Teh IPR* are more valuable than ever to keep the competitive advantatge Tech drivers * Intellectual Property Rights 4 Many customers are still on the edge of their journey to the Cloud Promise and value proposition is clear 17/07/2015 Enterprise roadblocks to move to Cloud Weight of legacy and fear of migration complexity ▶ Increase productivity ▶ Higher flexibility ▶ Accelerate the response to demands ▶ Elastic access to infrastructure resources ▶ Agility and virtual teams Complex Cloud market, Complex billing and management Localization of data and privacy to comply with regulations ▶ Reduce costs Enterprise-grade availability & Security missing in many offers ▶ … and it works! Reluctance to become prisoner of another technology silo 5 BIG DATA adoption: Drivers and Barriers 17/07/2015 BARRIERS DRIVERS • Immature technology • Adoption cost (storage ▶ Efficiency benefits outsourcing) ▶ Better services • Expertise and tech skills ▶ Innovation possibilities required to optimal operate ▶ Others are using it (successful cases) analyst and BI) ▶ Decrease of adoption cost privacy • Understand value (Data • Security and concerns on • Migration to cloud • Regulatory aspects • 6 BD: Data Storage & Processing 17/07/2015 Storage: (NoSQL concept elasticity and fault tolerance ) Key-value stores Column-oriented Documentoriented data Graph-oriented databases Voldermort (Linkedin), Membase Google BigTable, MongoDB Neo4j Cassandra (facebook), (10gen), Hbase (Yahoo, CouchDB Microsoft*) The choice of a solution depends on the strategy for the exploitation of Big Data chosen. Consistency models: Trade-off between consistency and availability! Processing: (Map reduce. Hadoop implementation) - We need an optimal ‘processing’ environment (cloud resources & configurations in private, public, federated, hybrid modes) - Reduce data transfer vs remote clouds - Map reduce designed for batch processes – so not suitable for real time! 7 Internet of Things Part 1 8,000 Units Installed Worldwide (millions) 7,000 6,000 5,000 Computers Handhelds Networking 4,000 Industrial/automotive Embedded 3,000 Household 2,000 1,000 0 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2016 2017 2018 2019 2020 Source: IDC Everything Network © 2012 IDC Jul-15 8 Internet of Things Part 2 1,400,000 Units Installed Worldwide (millions) 1,200,000 1,000,000 800,000 Sensors and Tags Everything Else 600,000 400,000 200,000 0 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2016 2017 2018 2019 2020 Source: IDC Everything Network © 2012 IDC Jul-15 9 17/07/2015 Information exploition ( immense dataset) ▶ Data generation rate and storage needs is rising faster than net bandwidth. ▶ Video-on-demand services occupied 30% of Internet bandwidth in December 2012. ▶ YouTube received 72 hours of new video every minute, which required 17 petabytes of new storage in 2012. ▶ Mobile devices will both consume and generate much of this data. By the end of 2012, mobile devices generated 25% of Internet traffic. ▶ According to Cisco, video will account for 86% of all wireless traffic by 2016. ▶ Mobile devices also generate lots of sensor data, such as GPS location data. Thus, they are the primary source of the machine-to-machine (M2M) traffic that comprises the Internet of Things. ▶ An IDC report forecasts that machinegenerated data will represent 42% of all data by 2020 (up from 11% in 2005). 10 17/07/2015 Western Europe Internet Devices 600,000,000 500,000,000 400,000,000 PC 300,000,000 Non-PC 200,000,000 100,000,000 Post-PC era 0 2010 2011 2012 2013 Source : IDC Information Society Index 2014 11 2015 Scenario (example): Smart Stadium Movement/capacity sensors 802.11 interface 17/07/2015 Crowd uploading content to social networks Server Media Distribution Media on-venue/internet distribution Bandwidth Intelligent waste management Public waste baskets monitor their fill level, frequency of use and defectiveness CDN Encoders Fingerprint capture Radio F IP cameras Sportmen recognition and 3D tracking >>> CPU Security Access capture personal informations and perform a verification in real time of that person against personal RFID badge of the enterprise. Mobile device Crowd enters the stadium Private/Public Cloud Tetra Security agents use a Public Security network 12 Content management Recommendation systems Augnmented reality 12 17/07/2015 Scenario (example): Smart Airport Mobile device & client app 802.11 interface Bluetooth Server Webcam Ethernet Weather sensors Online storage Operational DB Shopping facilities 13 13 Some hints when analysing the symbiosis (IoT, BD, Mobile, Clouds) 17/07/2015 ▶ Put business objectives and market cases at front (industry driven). ▶ Most IT organizations like to separate data, and cloud, and even assign them to different teams. However, it may be more productive to link them strategically. ▶ Big data and IoT segments will become more tightly coupled with Cloud as markets continue to progress. ▶ Don’t think that the fundamental technologies will merge at any point. Instead, look at the clear dependencies that should be considered when dealing with these technologies independently, and as a whole. ▶ Solve the lack of comprehensive vision and necessary skills to understand the interaction, impact and dependences of Big Data, IoT, Mobile and Clouds, all at the same time 14 17/07/2015 Some challenges • • • • • • • • Data scalability problem is not the same that Cloud scalability / elasticity problem (data assets are not VMs). Both strategies need to be aligned to deliver performance, reliability, consistency and availability. IoT related applications have non-virtualised parts (distributed sensors and agents) and it is necessary to study how to incorporate this in the Cloud Management layers (generally more centralized approach) Data management and sharing need better abstractions to be included in the Cloud programming models Strategies for the migration of huge volume of data to cloud Skills gaps in the intersection of Data management & Cloud delivery models Real time need vs BidData processing approach has limitations and impact on strategy for mobile clouds Hypervisor choice & resource type impacts on application data performance (not well understood yet). Need clouds specialization. Mobile access networks and context aware computing as the main mean to consume data. Offloading and dynamic bursting strategies needed at the edge of network. 15 ▶ For more information please contact: [email protected] Atos Av Diagonal 200 08017 Barcelona - Spain Atos, the Atos logo, Atos Consulting, Atos Worldline, Atos Sphere, Atos Cloud and Atos WorldGrid are registered trademarks of Atos SA. June 2011 © 2011 Atos. Confidential information owned by Atos, to be used by the recipient only. This document, or any part of it, may not be reproduced, copied, circulated and/or distributed nor quoted without prior written approval from Atos. 17/07/2015