#200 - Virtualisation, Containers, Serverless and Data

This week, Martin and Chris are joined by Jim Walker, VP of Product Marketing at Cockroach Labs. In a slight departure from a purely storage-based recording, this discussion follows the evolution of application packaging and deployment, from virtualisation to containers and serverless. The logical conclusion of any computing environment is to execute code and gradually, we are abstracting from the specifics of infrastructure and allowing the code to do the work.

How will serverless technologies standardise and implement non-functional requirements such as security and credentials management, data access and most important – portability? Will de-facto standards emerge or do we have to rely on standards’ bodies? We discuss all this and more.

Elapsed Time: 00:46:05

Timeline

00:00:00 – Intros
00:00:30 – Martin goes retro with CP/M and Wordstar
00:03:00 – The move to serverless is going to change the approach to application deployment
00:05:00 – How is the transition to serverless going?
00:07:10 – How does serverless compare to the mainframe days?
00:10:30 – Are security, storage, networking all sorted in serverless?
00:12:15 – Databases should be serverless, accessed by a SQL-based API
00:15:30 – What is the transition to serverless, are we doing it right?
00:20:30 – De-facto standards – are there any in serverless?
00:24:15 – Do the standards’ bodies actually work?
00:26:30 – Docker (the company) didn’t succeed but did popularise containers
00:31:45 – Do we need to think about data protocols rather than storage protocols?
00:36:00 – Physical factors like latency (and speed of light) affects distributed applications
00:38:45 – Why don’t we abstract data into unified service tiers?
00:41:00 – APIs still need a lot of work to make them generically consumable
00:44:00 – Billing and FinOps becomes important with Serverless
00:45:00 – Wrap Up

Related Podcasts & Blogs

Podcast: Play in new window | Download

(Visited 32 times, 1 visits today)