Come network and hear about how Elasticsearch can be leveraged for Distributed System Analysis as well as installed via Kubernetes. Food and refreshments will be provided on the second floor of WeWork at 745 Atlantic Avenue. A big thanks to WeWork South Station for offering their space and free beer.
Distributed System Analysis using a little TLC and ElasticSearch
A short presentation on our Distributed System Analysis efforts to facilitate indexing Telemetry, Log, and Configuration data to enable deeper analysis of complex distributed systems.
We all are working with ever increasing levels of complexity in our distributed systems as the need for more and more compute power and more storage grows. We describe the work we have done so far to build a warehouse for performance analysis, letting us separate visualization and reporting from collection and normalization, while providing a foundation for applying machine learning techniques. We'll walk through how we collect and normalize Telemetry data into ElasticSearch, along with Log data, and then how we collect and normalize configuration data. We show some of our work to enable Grafana to visualize the telemetry data, and demonstrate a performance analysis tool, pbench, which wraps benchmarks runs indexing them as discrete results sets, enabling a flexible web UI for visualization and reporting (nvd3 based).
This work is in its very early stages, so there will be lots of duct tape, band-aids, paper-clips, chewing gum, and bees-wax used to keep it all together. :)
Peter Portante is a Software Engineer at Red Hat.
Kubernetes: What, Why, and How with ELK
We will discuss what Kubernetes is, and is not, compared to previous deployment systems on physical/virtual machines, Ansible, and OpenStack and explain why Kubernetes is different.To understand how Kubernetes works, we will deploy ELK in a Pod to showcase some of the features Kubernetes provides.
Learnaholic, passionate about Distributed Systems and uptime, Myles Steinhauser is currently focused on ELK, Elixir, Containers and OpenStack.