As more and more services come into the cloud, the bar for service reliability, and the need for resources to monitor it, continues to increase. To put this challenge in context, consider the typical service provider’s SLA: 99.95% uptime translates to only 22 minutes of unplanned downtime per month, and 99.999% translates to less than just 1 minute per month. Proactive early detection of service issues is the key to achieving and maintaining service availability.
This tutorial session provides an overview of how to use Apache Spark and predictive modeling to improve your operational analytics, your capability to diagnose real-time service and device problems, and dramatically improve the service experience of your users. The content includes a reference architecture, and demonstrates how Spark can take massive quantities of data, both historical and real-time, and enable monitoring, analysis, and the prediction of problems before or shortly after they happen. This allows a materially proactive approach to your global services management.
This tutorial is sponsored by IBM
Romeo Kienzler holds a Master Degree in Information Systems with specialisation in Applied Statistics and Bioinformatics from the Swiss Federal Institute of Technology. He works for IBM Zurich as a chief data scientist and architect. His current research focus is on large-scale machine learning on resilient cloud infrastructures based on Apache Spark and Apache Flink on top of Docker Swarm.
Romeo Kienzler is a Member of the IBM Academy of Technology, the IBM Technical Expert Council and the IBM BigData BlackBelts team.
©2015, O’Reilly UK Ltd • (800) 889-8969 or (707) 827-7019 • Monday-Friday 7:30am-5pm PT • All trademarks and registered trademarks appearing on oreilly.com are the property of their respective owners. • firstname.lastname@example.org