Event details

All presentations will be delivered in English

REMEMBER TO REGISTER ON OUR BLOG! WITHOUT THIS YOU WON'T BE ABLE TO ENTER THE MEETUP 

https://allegro.tech/events/  (PLEASE USE REGISTER BUTTON - few minutes after registration you will receive confirmation e-mail together with a ticket) 

Schedule

18:00 - Registration starts (+ Networking & pizza) 

18:30 - Debug-Companion Tomasz Ziarko

19:30 - Container 2.0 (Johannes Unterstein)

Debug-Companion

Switching from VMs to Mesos has a lot of benefits for developers, but also one huge drawback: they are no longer in full control of the environment. For JVM this means losing access to tools like jstat, jmap, or perf. Giving full access to all containers on Mesos is too dangerous while asking ops each time a developer wants to create a heapdump contradicts all DevOps principles.

This is why we decided to take a different approach and give back the power to the developers. During this talk we will present different problems with debugging apps on our Mesos cluster running 400+ services operated by 40 development teams. We will cover security and stability concerns and present our solution: debug-companion.

Tomasz Ziarko

Programmer/Engineer with 7 years experience, on a daily basis dealing with cloud infrastructure in Allegro.pl, specifically with Mesos and Openstack. A supporter of open source software.

Container 2.0

In the past year, we migrated our applications to containers to isolate our applications and made them easily shippable. But now we are struggling running stateful applications to make them scalable and fault tolerant. And we run our big data and fast data frameworks in separate clusters because it is not super easy to integrate this into our current container stack. This means that our cluster utilization is not optimal and we resign scaling effects comprehensive containers and big data frameworks. Florian Leibert (Mesosphere, CEO) calls this phenomenon Container 2.0 and we will discuss this topic in this session.
In this session, we will look at current container usages and show the actual problem. After this, we will see how DC/OS and Apache Mesos delivers Container 2.0 to optimize cluster resource sharing for all parts of your application.

Johannes Unterstein (@unterstein)

organizes the Java user group in his hometown Kassel, teaches at DHBW Stuttgart and works as a distributed applications engineer at Mesosphere. He is working at the Marathon container orchestration framework to provide Twitter scale for everyone. In his previous life, he worked in java projects for large German customers. Besides his fascination for container orchestration, he loves to talk about elasticsearch and Neo4j.