Leveraging software optimization to reduce cloud footprint
summary
Cloud object storage sits in exabytes and companies spend billions a year to store and retrieve it; applying classic file-system tricks — compression, deduplication and encryption — to object stores reduces costs but hits concrete hurdles around overwrites, versioning and encrypted blobs. The talk lays out when client-side encryption vs server-side encryption permits dedupe, why online deduplication is required for encrypted objects, practical tradeoffs (CPU, latency, metadata), and cites Netflix/Box-style deployments that combine techniques to save millions in real deployments.
AI-written summary · about
chapters (6)
- 0:00Introduction to cloud storage
- 2:40Compression basics explained
- 5:20Data deduplication techniques
- 8:14Encryption for security
- 10:36Online vs offline replication
- 13:49Q&A and summary
related talks
14mInftyDedup: Scalable and Cost-Effective Cloud Tiering with Deduplication
30mHow to Facilitate the Cloud Foundry Adoption with Object Storage - Denis Jannot, Dell EMC
49mOptimizing Data Storage Costs in Object Storage
19mObject Compaction in Cloud for High Yield
30mCost optimization in Amazon S3, Amazon EBS, and beyond
36mSustainable cloud infrastructure: from ideation to deployment
33mThe elephant in the cloud: Strategies of cloud cost optimisation
20mSoftware Performance Optimization and Scaling