Practical evaluation of the Lasp programming model at large scale

Abstract

Programming models for building large-scale distributed applications assist the developer in reasoning about consistency and distribution. However, many of the programming models for weak consistency, which promise the largest scalability gains, have little in the way of evaluation to demonstrate the promised scalability. We present an experience report on the implementation and largescale evaluation of one of these models, Lasp, originally presented at PPDP '15, which provides a declarative, functional programming style for distributed applications. We demonstrate the scalability of Lasp's prototype runtime implementation up to 1024 nodes in the Amazon cloud computing environment. It achieves high scalability by uniquely combining hybrid gossip with a programming model based on convergent computation. We report on the engineering challenges of this implementation and its evaluation, speci cally related to operating research prototypes in a production cloud environment. ACM Reference format:C. Meiklejohn et al.of demonstrating the scalability of the Lasp model. Our experience report substantiates that empirically validating scalability is nontrivial, regardless of the programming model.

Conflict-Free Replicated Relations for Multi-Synchronous Database Management at Edge

“…To study the performance of CRR, we implemented an advertisement-counter application [18,21] in Crecto. Advertisements are displayed on edge devices (such as mobile phones) according to some vendor contract, even when the devices are not online.…”

“…It requires a central server in the cloud to disseminate operations and the data model is restricted to the MangoDB 4 model. Lasp [18,21] is a programming model for large-scale distributed programming. One of its key features is to allow local copies of replicated state to change during periods without network connectivity.…”

Secure Scuttlebutt

“…Selective hearing was developed in a more traditional research approach, so it glosses over some of the difficult problems encountered in the "real world" such as user onboarding and byzantine peers. Their "practical large scale evaluation" [27] consists of 1000 well-controlled nodes running a toy application in the cloud, whereas SSB with its roughly 10000 users is more battle-proven.…”

Achlys: Towards a Framework for Distributed Storage and Generic Computing Applications for Wireless IoT Edge Networks with Lasp on GRiSP

“…We now present a more detailed description of the Achlys system, an Erlang 2 implementation of a framework that combines the power of δ-CRDTs [22] in the Lasp store [23], [24], the Partisan communication component [25], and the GRiSP Runtime software. It provides application developers a way to build resilient distributed edge IoT applications.…”

“…Lasp is a key part of the Achlys framework for both storage and computation. It provides both replicated data and computation, and guarantees that values will eventually converge on all nodes [23], [24]. Since our GRiSP boards run Erlang directly on bare metal, Lasp is a suitable option for consistency as it runs directly in Erlang [24].…”