S3DMap: A visual storage map for prefix-level cost optimization methodology

Imagine having a layered map for S3 storage, allowing you to interactively visualize objects by their prefix/class/cost/access/ data transfer/etc., in order to compose an optimal storage class strategy.

In this talk, I will introduce an advanced, prefix-oriented methodology (and a GIS-like tool) for optimizing Amazon S3 storage – a step forward from traditional bucket-oriented approaches. This strategy, supported by a novel open-source tool – S3DMap, emphasizes the importance of detailed prefix-level analysis in reducing S3 costs. I will discuss how focusing on prefixes, rather than entire buckets, enables a more granular and effective optimization of S3 storage and usage. The talk will showcase how S3DMap (inspired by SpaceMonger from the 2000's) assists in this approach, serving as a layered map at the prefix/object level. Attendees will leave with a clear understanding of the benefits of this approach and how to implement this methodology for enhanced S3 efficiency and cost savings.