Unlocking Storage Efficiency: Diagnosing & Optimizing Performance in Scaling Infra
summary
Work on OpenShift’s container image storage diagnoses S3 multipart uploads that silently create incomplete blobs; about 3,000 flaky uploads with five retries escalate to ~15,000 needless operations. Fix implements explicit verify each blob hash after multipart upload and aborts on mismatch (PR #43 to Docker distribution), rolled out in stages. Measured impact shows 30–40% pull latency reduction, retries fall to near zero, blob-mismatch errors vanish and S3 egress costs drop.
AI-written summary · about
chapters (10)
- 0:00Hidden cloud costs
- 2:07CI inefficiencies
- 6:27OpenShift image storage failures
- 11:51Storage observability
- 14:19Multipart uploads validation
- 18:37Speed reliability cost balance
- 20:05Storage issue diagnosis
- 22:15Storage performance best practices
- 24:04Key takeaways
- 25:10Q&A
related talks
44mOptimizing storage price and performance with Amazon S3
25mOptimizing Container Registry Storage: Introducing Blob De-Duplication for Accurate Sizing and Cost
23mOn-demand Container Loading in AWS Lambda
25mStreamlined Efficiency: Unshackling Kubernetes Image Volumes for Rapid AI Model...
10mCost Optimization in Cloud Foundry Through Economies of Scale
42mCloud and Mobile App Performance
17mBoosting Cost Efficiency by Cost Aware Architecture
18mImproving your Amazon S3 security with cost-effective practices