|
|
|
|
|
by Pyxl101
3471 days ago
|
|
I also have been finding this difficult to understand. It doesn't help that the Wikipedia article for NFV reads like vendor marketing copy: > NFV has proven a popular standard even in its infancy. Its immediate applications are numerous, such as virtualization of mobile base stations, platform as a service (PaaS), content delivery networks (CDN), fixed access and home environments. The potential benefits of NFV is anticipated to be significant. Sounds like it can do everything! By comparison, the paper is apparently more self-reflective and says: > despite[] promised advances, there has been little progress towards large-scale NF deployments Skimming the paper and comparing it to MapReduce, it seems to provide a MR-like programming model tailored to packet processing. In section 3.1 they describe programming control flow abstractions: group by, shuffle, merge. These should be familiar to users of distributed compute frameworks like Hadoop. NFV seems to be about implementing this style of compute while processing packets at network line rate. |
|