Tachyon meetup

Tachyon is a memory-centric distributed storage system.
Today it is used mostly as a storage system for Spark's serialized blocks, in order to get predictable performance over S3, and for other interesting applications. In my opinion, it has great potential, especially in the cloud. I would say that as Spark is preempting MapReduce, Tachyon might preempt HDFS in the cloud.
In this meetup we are going to go over Tachyon's architecture, examine concepts of lineage, tiered storage, Underlying file system.
We will see examples of Tachyon usage - by Spark and by ImpalaToGo, and maybe even more.
If you have experience with Tachyon and are willing to present real life cases - please let me know and I will share the stage.
This meetup is kindly hosted by Verint. They will give us overview of their big data usage. I am sure they have interesting challenges to share.