Internship: Fast System Deployment on Cluster for Distributed System Experiment with NixOS
- Level: Master Level Research Internship (M2 & M1)
- Location: University Grenoble Alpes Campus, Saint Martin d’Heres (close to Grenoble)
- Duration: At least 4 months, possibility to pursue as a PhD.
- Contact: Olivier.Richard@inria.fr
- Incomes: Gratifications de stage (about 500 euros/month)
- Period: 2020-2021
Work
Fast system deployment is a key issue to speedup experiments of distributed systems.
Grid’5000 platform [1] provides the Kadeploy tool [2][3] which uses efficient system image broadcasting but requires several costly (in time) reboots of involved machine. We want to explore the use of Nix [4][5] a Functional Package Management [4] and some system tricks to provide fast deployment system image alternatives.
References
[2] Emmanuel Jeanvoine, Luc Sarzyniec, Lucas Nussbaum. Kadeploy3: Efficient and Scalable Operating System Provisioning for Clusters. USENIX Association, USENIX Association, 2013, 38 (1), pp.38-44
https://hal.inria.fr/hal-00909111/
[3] Luc Sarzyniec, Sébastien Badia, Emmanuel Jeanvoine, Lucas Nussbaum. Scalability Testing of the Kadeploy Cluster Deployment System using Virtual Machines on Grid’5000. SCALE Challenge 2012, held in conjunction with CCGrid’2012, May 2012, Ottawa, Canada.
https://hal.inria.fr/hal-00700962/
[5] Nix: A Safe and Policy-Free System for Software Deployment: LISA 2004. https://nixos.org/~eelco/pubs/nspfssd-lisa2004-final.pdf
