[Nfvrg] IETF NFV Research Group (NFVRG) interim meeting -- Re-thinking VNF Architectures towards a Cloud-native Deployment

[Ramki Krishnan <[e-mail]>]

Dear All,

We would like to invite you to an NFVRG interim meeting, details are below. Please fill the doodle poll (https://doodle.com/poll/bt4nmwqbvdnxy743). In the doodle comment field, please indicate if you are planning to attend in person or join remotely.

Thanks,
NFVRG Co-chairs (Ramki Krishnan & Diego Lopez)

Interim Meeting Theme:             Re-thinking VNF Architectures towards a Cloud-native Deployment

Interim Meeting Time:                 10:00am to 12:00pm Sept. 18th Monday

Interim Meeting Location:
Conf - Creekside C - Darwin (CSC1139)(Seats 72, Projector)
900 Arastradero Road
Palo Alto, California, 94304
VMware campus map -- https://www.vmware.com/files/pdf/vmw-directions-to-vmware.pdf

Webex link: https://vmware.webex.com/mw3200/mywebex/cmr/cmr.do?siteurl=vmware&AT=meet&username=ramkik

Presentations

  *   A New Approach to Network Functions, Aurojit Panda                               40 minutes (30 minutes talk + 10 minutes Q&A)
  *   Building a better network through disaggregation, Eric Keller                   40 minutes (30 minutes talk + 10 minutes Q&A)

Brainstorming Discussion                                                                                                  30 minutes

  *   Progressing this topic in NFVRG leveraging current work on lightweight virtualization technologies
  *   Progressing this topic in Open Source Efforts such as ONAP, OSM, OPNFV, FD.io etc.

Abstract - A New Approach to Network Functions
Modern networks do far more than  just deliver packets, and provide network functions -- including  firewalls, caches, and WAN optimizers - that are crucial for scaling networks, ensuring security and enabling new applications.  Network functions were traditionally implemented using dedicated hardware middleboxes, but in recent years they are increasingly being deployed as VMs on commodity servers. While many herald this move towards network function virtualization (NFV) as a great step forward, I argue that accepted virtualization techniques are ill-suited to network functions. In this talk I describe NetBricks - a new approach to building and running virtualized network functions that speeds development and increases performance.

Abstract - Building a better network through disaggregation
To improve performance, security, and reliability, network practitioners have, over time, moved away from the principle of a stateless network and added stateful processing to the network with devices such as firewalls, load balancers, and intrusion detection systems. In doing so, networks have become increasingly complex and brittle, because the state held in these devices (such as the connection tracking information in a firewall) is needed to process the traffic. The conventional approach forces practitioners to configure or architect the network to get the right traffic to the right (physical or virtual) appliance (i.e., where the relevant state is), and introduce costly, and sometimes ineffective, mechanisms to back up state (e.g., to recover from failures).  In a world where agility is increasingly important, a new approach is needed.

In this talk, we present our a network architecture based on disaggregated network functions. Our foundational work breaks the underlying assumption that state needs to be tightly coupled to a specific device, the state is maintained separately and the network functions can access that state from anywhere and at any time through a well-defined interface - creating a highly flexible network.  After years of research, we proved this architecture viable (publishing the results at NSDI), and now we are commercializing at Stateless, Inc.  In this talk we will present the background and technical details of this disaggregated architecture, discuss the challenges we are currently working on, and the use cases driving the commercial adoption.

Aurojit Panda Bio
Aurojit Panda is currently a software engineer at Nefeli Networks, and starting in Fall 2018 an assistant professor in computer science at the Courant Institute in New York University. He received his PhD in Computer Science from the University of California, Berkeley, where he was advised by Scott Shenker.  His work spans programming languages, networking and systems, and his recent work has investigated network verification, consensus algorithms in software defined networks, and frameworks for building network functions.

Eric Keller Bio
Eric Keller is co-founder and CTO of Stateless, Inc. a spin-off from the University of Colorado, Boulder where he is also an Assistant Professor. Dr. Keller joined CU after receiving a PhD from Princeton, and generally works on networking and security, with special interest in building and leveraging programmable infrastructures (SDN, virtualization, etc.).  Stateless was founded to commercialize the research of Dr. Keller and his first PhD student (and now CEO of Stateless), Murad Kablan.  The company is a recent graduate of Techstars Boulder, has grown to a team of 9, received an SBIR award, and has two deployments in progress.