VMware sponsored this podcast.
The technology industry is undergoing a data revolution in which the unlimited storage and compute power of the cloud is changing how data is stored, processed and managed.
Alluxio is a platform for data orchestration that aims to simplify and standardize how data is managed across different types of infrastructure by creating a layer of abstraction between the storage and application layers.
The Alluxio orchestrator virtualizes data and allows applications to access it in a way that’s compute, storage and cloud-agnostic. It’s a platform designed to eliminate data silos and make data readily available and performant for developers.
Haoyuan (H.Y.) Li, CTO and founder of Alluxio, is a co-creator of the Apache Spark streaming library and built Alluxio as an open source virtual distributed file system for a computer science Ph.D. project at Berkeley. Li is now building a startup, which currently boasts 40 employees and several large enterprise customers. The company offers enterprise features and support on top of its open source project.
“There are so many data silos which [are] caused by many storage systems in the industry,” Li said. “In fact, every five to 10 years there’s another generation of storage systems being created. These data silo issues won’t go away and it won’t be solved by a new storage system… We looked at this problem and took a dramatically different approach than the rest of the data industry.”
On this episode of the Makers podcast, The New Stack editorial and marketing director Libby Clark talks with Li about what data orchestration is, where it fits into the cloud native stack and how it can make on-premises data more accessible in the cloud, and vice versa. Making this connection between the application and data stores on-premises and in the cloud is a key challenge in providing artificial intelligence and realtime analytics capabilities to the enterprise.
In this Edition:
1:39: Could you start by talking for a minute about the origin story of Alluxio?
6:40: How do you view this as a company, and creating a company on top of that sort of industry-wide common layer?
16:00: How is it that you are able to improve that performance?
19:28: So what is the scenario for Kubernetes? How do you sit in relation to a storage orchestrator with Kubernetes? Are teams then writing a bunch of YAML files to do the provisioning?
22:48: So what are some of the drawbacks or compromises to this approach?
25:55: How are you making sure that those policies and access points are secure?
Feature image via Pixabay.