Talks (order to be decided):

• Percolate on moving from Mysql to ES

• ES 5 ingest node (by Suyog Rao from Elastic)

• Idempotent systemd log aggregation (lightning)

The Percolate Analytics team will explain why they moved to Elasticsearch for the next generation of their Analytics platform, which is currently in Beta. The team will describe how the data modeling changed with the move to Elasticsearch, integration with Apache Kafka, production monitoring, and lessons learned along the way. They will also discuss future ideas for the system.

Food and beverages will be provided.

Presenters:

• Vinh Nguyen, Senior Engineer, Analytics

• Yannick Dawant, Software Engineer, Analytics

There is a blog post about the project (see last section, titled "analytics"): https://blog.percolate.com/2016/09/percolate-new-features-marketing-visibility-coordination/

Elastic 5.0 has new features across the stack for every one of our users. In this talk, I'll present highlights in 5.0 across our ingestion products. I will deep dive into ingest node, a new feature that allows users to perform light-weight data transforms directly in Elasticsearch.

Presenter: Suyog is a team lead at Elastic. He primarily develops for the Logstash project, supports customers, and trains users on the Elastic Stack. Previously, he worked on a high-throughput, low-latency infrastructure for ingesting and analyzing terabytes of log data using Elasticsearch, Apache Kafka, and Storm.

Mik Kocikowski will show how to idempotently aggregate systemd / journald logs with 'journalbeatlite': https://github.com/pokitdok/journalbeatlite