STI-BT: A Scalable Transactional Index

Abstract

In this article we present STI-BT, a highly scalable, transactional index for Distributed Key-Value (DKV) stores. STI-BT is organized as a distributed B+Tree and adopts an innovative design that allows to achieve high efficiency in large-scale, elastic DKV stores. We have implemented STI-BT on top of a mainstream open-source DKV store and deployed it on a public cloud infrastructure. Our extensive experimental study reveals the efficiency of our solution with demonstrable scalability in a cluster of 100 commodity machines, and speed ups with respect to state of the art solutions of up to 5.4x.

Optimizing hyperspace hashing via analytical modelling and adaptation

“…Unlike hyperspace hashing, these techniques do not determine the placement of replicas of data using multidimensional hashing schemes. Conversely, these solutions assume that the placement of data is governed by an orthogonal placement policy (e.g., consistent hashing) and build indexes over secondary attributes using distributed tree-like data structures [2,28,25,14]. Despite being designed to maximize efficiency and scalability, also these approaches clearly incur costs to maintain and query distributed indexes, which vary also depending on the consistency semantics that they ensure (ranging from eventual consistency [10] to classic 1-copy serializability [3] and including intermediate consistency semantics [1]).…”