Some Scientific Challenges in the Cloud - Roberto Di Cosmo
•
1 like•880 views
The document discusses some scientific challenges in cloud computing. It defines the three layers of cloud computing: Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS). For IaaS, key challenges include optimizing resource sharing and placement of virtual machines, as well as security issues like avoiding escapes. For PaaS, ensuring flexibility without vendor lock-in is difficult. Transversal challenges that apply across layers include security, data ownership, efficient implementation, and interoperability between services and vendors. Overall, the document argues that more work is still needed to address challenges in cloud computing.
Some Scientific Challenges in the Cloud - Roberto Di Cosmo
- 1. Some Scientific Challenges in the Cloud Roberto Di Cosmo University Paris Diderot, and INRIA May 20, 2010 GTLL Paris Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 1/5
- 2. Cloud Computing: basic definitions Three levels: Layer Main target Abstraction Software as a Service (SaaS) Final users Application Platform as a Service (PaaS) App Developers System Infrastructure as a Service (IaaS) Platform Developers HW Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 2/5
- 3. Cloud Computing: basic definitions Three levels: Layer Main target Abstraction Software as a Service (SaaS) Final users Application Platform as a Service (PaaS) App Developers System Infrastructure as a Service (IaaS) Platform Developers HW Key concepts: virtualization Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 2/5
- 4. Cloud Computing: basic definitions Three levels: Layer Main target Abstraction Software as a Service (SaaS) Final users Application Platform as a Service (PaaS) App Developers System Infrastructure as a Service (IaaS) Platform Developers HW Key concepts: virtualization + flexibility Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 2/5
- 5. Cloud Computing: basic definitions Three levels: Layer Main target Abstraction Software as a Service (SaaS) Final users Application Platform as a Service (PaaS) App Developers System Infrastructure as a Service (IaaS) Platform Developers HW Key concepts: virtualization + flexibility + automation Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 2/5
- 6. Cloud Computing: basic definitions Three levels: Layer Main target Abstraction Software as a Service (SaaS) Final users Application Platform as a Service (PaaS) App Developers System Infrastructure as a Service (IaaS) Platform Developers HW Key concepts: virtualization + flexibility + automation Erich Clementi, IBM’s Cloud Initiative head says it well: Many people equate cloud computing to virtualization. It is not virtualization. To get the value you need standardization and automation on top of that virtualization. Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 2/5
- 7. Cloud Computing: basic definitions Three levels: Layer Main target Abstraction Software as a Service (SaaS) Final users Application Platform as a Service (PaaS) App Developers System Infrastructure as a Service (IaaS) Platform Developers HW Key concepts: virtualization + flexibility + automation Erich Clementi, IBM’s Cloud Initiative head says it well: Many people equate cloud computing to virtualization. It is not virtualization. To get the value you need standardization and automation on top of that virtualization. Let’s sketch some scientific challenges at each level. Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 2/5
- 8. Challenges in IaaS Optimization of resources: Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 3/5
- 9. Challenges in IaaS Optimization of resources: share common data in disk and memory (see Decentralized Deduplication in SAN Cluster File Systems, Clements et al., and Satori: Enlightened page sharing, Milos et al. USENIX ’09) optimal VM placement Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 3/5
- 10. Challenges in IaaS Optimization of resources: share common data in disk and memory (see Decentralized Deduplication in SAN Cluster File Systems, Clements et al., and Satori: Enlightened page sharing, Milos et al. USENIX ’09) optimal VM placement Security: Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 3/5
- 11. Challenges in IaaS Optimization of resources: share common data in disk and memory (see Decentralized Deduplication in SAN Cluster File Systems, Clements et al., and Satori: Enlightened page sharing, Milos et al. USENIX ’09) optimal VM placement Security: avoiding escapes Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 3/5
- 12. Challenges in IaaS Optimization of resources: share common data in disk and memory (see Decentralized Deduplication in SAN Cluster File Systems, Clements et al., and Satori: Enlightened page sharing, Milos et al. USENIX ’09) optimal VM placement Security: avoiding escapes avoiding mapping of the physical infrastructure, and its exploitation http://cseweb.ucsd.edu/~hovav/dist/cloudsec.pdf Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 3/5
- 13. Challenges in PaaS Ensuring flexibility, avoiding vendor lock-in, is not easy. Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 4/5
- 14. Challenges in PaaS Ensuring flexibility, avoiding vendor lock-in, is not easy. Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 4/5
- 15. Challenges in PaaS Ensuring flexibility, avoiding vendor lock-in, is not easy. N.B.: (re)writing applications for a platform/middleware is not neutral. Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 4/5
- 16. Transversal Challenges Several challenges are transversal: security: who can break (into) the system? Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 5/5
- 17. Transversal Challenges Several challenges are transversal: security: who can break (into) the system? data ownership: who can access my data? Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 5/5
- 18. Transversal Challenges Several challenges are transversal: security: who can break (into) the system? data ownership: who can access my data? please do not simply rely on contracts! Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 5/5
- 19. Transversal Challenges Several challenges are transversal: security: who can break (into) the system? data ownership: who can access my data? please do not simply rely on contracts! efficient implementation: abstraction is nice, but can it accomodate optimal implementation? Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 5/5
- 20. Transversal Challenges Several challenges are transversal: security: who can break (into) the system? data ownership: who can access my data? please do not simply rely on contracts! efficient implementation: abstraction is nice, but can it accomodate optimal implementation? interoperability: SaaS and PaaS try to hide the complexity of the lower layer, but at the price of vendor or technology lock-in Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 5/5
- 21. Transversal Challenges Several challenges are transversal: security: who can break (into) the system? data ownership: who can access my data? please do not simply rely on contracts! efficient implementation: abstraction is nice, but can it accomodate optimal implementation? interoperability: SaaS and PaaS try to hide the complexity of the lower layer, but at the price of vendor or technology lock-in Bottomline: there is a lot of work ahead. Roberto Di Cosmo () Scientific challenges in the Cloud May 20 2010 / Paris 5/5