Image of Rocky Jaiswal & Michal Olah

Rocky Jaiswal & Michal Olah

Rocky is a software developer with over a decade of experience in software design and programming. He enjoys coding in Ruby, JavaScript, Scala and Go. He loves working on open-source projects and tinkers with technology in his free time. He has been a speaker at AgileNCR 2010, Agile Tours 2010, IndicThreads Conference on Cloud Computing 2011, Ruby Conf India 2012&2013 and IndicThreads Software Development Conference 2012&2013. He is currently working as a programmer with crealytics GmbH, Berlin. His blogs / videos can be found on http://rockyj.in

Michal is a software developer with more than 7 years of IT experience, providing software design, development and implementation of internet/intranet applications for data management and data analysis projects. His scope of expertise includes web application development built upon existing-standards (Ruby on Rails, J2EE, JSP, JavaScript, MS Visual C#, Sass, Haml). He also has wide experience with project development, research and collaborations on various design and planning levels.

“Scaling Ruby applications with Mesos and Docker”

Scaling Ruby applications is hard. While PaaS solutions like Heroku provide such benefits the costs may be quite high. Most of us our looking for the flexibility of our own bare-metal servers with the scalability of a PaaS.
Docker meanwhile has also provided a great way to isolate or containerize our application. While Docker is great for distribution and deployment, we wonder can we also use it to scale our applications, i.e bringing containers up/down on demand.
Apache Mesos is one such technology we can leverage to add scale to our infrastructure. Instead of dealing with multiple machines we deal with our cluster as a single unit. We allocate resources from this single unit to Docker containers, making it very easy to add/remove resources.
In this talk, we will see a live demo of a Mesos cluster and what it takes to set it up. We will also Dockerize a simple JRuby service and add / scale it to our Mesos setup. We will also learn the advantages and pitfalls of this setup.