Mojave: A Recommendation System for Software Upgrades


Rekha Bachwani, Rutgers University; Olivier Crameri, EPFL; Ricardo Bianchini, Rutgers University; Willy Zwaenepoel, EPFL


Software upgrades are frequent. Unfortunately, many of the upgrades either fail or misbehave. We argue that many of these failures can be avoided for new users of each upgrade by exploiting the characteristics of the upgrade and feedback from the users that have already installed it. To demonstrate that this can be achieved, we build Mojave, the first recommendation system for software upgrades. Mojave leverages data from the existing and new users, machine learning, and static and dynamic source analyses. For each new user, Mojave computes the likelihood that the upgrade will fail for him/her. Based on this value, Mojave recommends for or against the upgrade. We evaluate Mojave for two real upgrade problems with the OpenSSH suite. Initial results show that it provides accurate recommendations.

Open Access Media

USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are posted after the event are also free and open to everyone. Support USENIX and our commitment to Open Access.

@inproceedings {179437,
author = {Rekha Bachwani and Olivier Crameri and Ricardo Bianchini},
title = {Mojave: A Recommendation System for Software Upgrades},
year = {Submitted},
url = {},
publisher = {{USENIX} Association},

Presentation Audio