Future Data Platforms and PaaS Services
summary
Conference argues current PaaS data stacks use many specialized systems—Cassandra, Redis, Elasticsearch, Solr, Riak/React, Hadoop—forcing separate ops, custom bridges and painful syncs; adding nodes and removing them is not seamless. Proposes unified storage that covers graph/document/key‑value/search and automates node scaling, reducing expert silos. Estimates relational workloads are ~15% and a single unified product would solve roughly 80% of NoSQL needs. Highlights concrete costs like sixfold replication causing 6× storage and backup overhead.
AI-written summary · about
chapters (8)
- 0:00Introduction to future data platforms
- 1:26What is a data platform
- 3:30Storage types in PaaS
- 9:17Readymade vs made-to-order services
- 14:00Challenges of today's storage services
- 17:58Vision for unified storage
- 23:15Future of vendor integration
- 26:03Q&A on data platform trends
related talks
05mAddressing Operational Challenges...
25mPlatform Engineering for Modern Data Infrastructure
34mData Storage: NoSQL Toasters and a Cloud of Kitchen Sinks
43mSoftware Quality As It Relates To Data
41mModern Data Stack: Optimising and Scaling Data in a Tech Company
15mWhen is it time to modernize your AWS database platform?
41mThe New Generation Of Data Stores
33mPredicting the Future of Distributed Systems