Pravega, a new storage abstraction data streams

Streaming Analytics has become an indispensable tool for any agile organization. While the rapid evolution of streaming tools has provided the ability to react to incoming data instantly, it has also introduced several engineering challenges. Data need to be duplicated to traditional storage systems for historical processing and archival, where a different set of tools are used for analyzing it. Pravega, an open source technology, aims to revolutionize this space by providing the first Streaming Storage solution that enables both Streaming and Batch Analytics using a single set of tools, all without unnecessary data replication. Designed from the ground up with exactly-once semantics, elastic stream scaling, and transactional support, Pravega hopes to revolutionize the streaming data platforms by providing stream as a data storage primitive.

Speaker Details

Srikanth Satya is the Vice President of Engineering for Unstructured Data Storage, responsible for spearheading the engineering of both the Object Storage Product – ECS, and a modern data analytics product (code-named Nautilus) which uses Pravega Streaming Storage (www.pravega.io (opens in new tab)). Pravega reimagines storage for the stream processing world, establishes stream as fundamental storage primitive and was recently open sourced by his team. Nautilus hopes to challenge the current status quo in the industry and unifies batch and stream processing for big data.

Srikanth joined EMC in 2015 and before that, he started several pioneering large-scale public cloud services at Amazon and Microsoft. Outside of work, Srikanth enjoys playing cricket, heading a non-profit that is actively working on inculcating innovative spirit in early childhood education and losing his voice and mind over Seattle Seahawks football games.

Andrei Paduroiu is among the original group of developers who started the Pravega project and is currently a core contributor, owning the data plane server-side aspects of it – the Segment Store. He holds an MS Degree from Northeastern University and a BS Degree from Worcester Polytechnic Institute. Previously, Andrei held software engineering positions with Microsoft and Vistaprint. Andrei’s interests include distributed systems, search engines, and machine learning.

Tom Kaitchuck is a Technical Staff Engineer at Dell EMC. Tom has spent his entire career working in messaging and streaming storage. Before joining Dell EMC, Tom worked at Google on Google Pipelines, Google PubSub, and AppEngine, and at Amazon where he designed the storage systems for AWS’s Simple Queue Service (SWF), Simple Workflow Service (SWF), and Amazon’s Messaging Platform (AMP). Tom is currently working on Pravega, a new open source Streaming Storage platform and is excited to share details on how Pravega provides users a low-latency, strongly consistent, way to durably store infinite streams of data leveraging our existing storage technologies.