^ Top

NANOG 65 Agenda

All times will be listed in Eastern Time.

**Agenda is subject to change.**

 

The NANOG 65 Agenda icalendar feed is now available! 
Download the .ics file and import into your calendar, or subscribe to the URL:

 http://www.nanog.org/sites/default/files/ical/nanog.ics

Floor Plans 

Download the Guidebook app!

If you already have the app, you can search for NANOG 65 and download the guide to your mobile device!

Sunday, October 4 2015
Time/Webcast:Room:Topic/Abstract:Presenter/Sponsor:Presentation Files:
4:00pm - 6:00pmConvention FloorRegistration
6:00pm - 8:00pmOffsite - Moxie'sSocial
Additional information
Sponsors:
Monday, October 5 2015
Time/Webcast:Room:Topic/Abstract:Presenter/Sponsor:Presentation Files:
8:30am - 11:00amHochelaga 1-4Extended BreakfastSponsors:
8:30am - 5:00pmConvention FloorRegistration
10:00am - 10:30am

Le Grand Salon & Marquette

Conference Opening

Welcome and orientation from Host and Connectivity Sponsors

View full abstract page.
Speakers:

  • Philippe Couture, Videotron
  • Philippe Couture, Eng., M.Sc., MBA Director, IP/MPLS Backbone, Research & Development, Engineering, Vidéotron G.P. Philippe has been involved in data network engineering for the past 20 years with various service providers and network equipment vendors. He joined VidÈotron, the 3rd largest Canadian MSO, in 2004 and has overall development responsibility for backbone architecture and services that support Internet, mobile, telephony, video and business services.
  • Christian S. Tacit, TorIX.
  • Tony Tauber, Comcast
  • In his role as Distinguished Engineer at Comcast, Tony focuses on Backbone and Core network architecture and engineering with particular attention to measurement, manageability, and automation. He also partners with the research and education communities on projects and currently chairs the NANOG Program Committee. In the past Tony held senior network engineering positions at BBN, GTE Internetworking, Genuity, Level3, and MIT Lincoln Lab as well as served as co-chair of the Routing Protocol Security working group in the IETF.

  • Clinton Work, TELUS
  • Sr Design Specialist with TELUS since 2001 and based out of Calgary, AB. My focus at TELUS is on MPLS/IP core network support and engineering. I was involved in building the TELUS MPLS core back in 2003 and all the services that we have layered on top since. I have extensive experience with Internet peering, MPLS, ISIS/BGP, L3VPNs, VOIP, and LTE/HSPA+. More recently I have focused on CDN cache server deployments and the TELUS IPV6 program for consumer services (DSL/GPON).
pdfConference Opening(PDF)
youtubeConference Opening
10:30am - 11:15am

Le Grand Salon & Marquette

NANOG 65 Keynote Address

Jack Waters, CTO of Level3 Communications will provide remarks and a vision for the future of our industry.

View full abstract page.
Speakers:

  • Jack Waters, Level3 Communications
  • Jack Waters is the chief technology officer for Level 3 Communications, focused on global network technology, architecture, engineering, process and security that are critical for Level 3 to deliver the solutions its customers need. Mr. Waters joined Level 3 in 1997 and has held numerous roles with the company spanning operations and engineering. Prior to his current role, Mr. Waters led Level 3’s global operations team. Before joining Level 3, Mr. Waters served as an executive staff member for MCI Communications, Inc. with responsibility for network architecture, design and implementation. In that role, he was also an original member of the team responsible for InternetMCI, the company’s initial Internet service. Prior to MCI, Mr. Waters worked as director of engineering and operations for SURAnet, the Southeastern University Research and Academic Network.
pdfNANOG 65 Keynote Address(PDF)
youtubeNANOG 65 Keynote Address
11:15am - 12:00pm

Le Grand Salon & Marquette

Your Bitcoins or Your Site: An Analysis of the DDoS for Bitcoins (DD4BC) DDoS Extortion Campaign.

In this presentation, we will discuss details of the ongoing DDoS for Bitcoins (DD4BC) DDoS extortion campaign, which has targeted multiple organizations in North America, Europe, Asia, Australia and New Zealand. The attack methodologies, evolution of the attack campaign, and discussion of the nature of the attacker(s) will be covered, along with a discussion of successful DDoS mitigation techniques which have proven effective against these attacks.

View full abstract page.
Speakers:

  • Roland Dobbins, Arbor Networks
  • Roland Dobbins has 30 years of operational experience in the service provider (SP) and large enterprise arenas, designing, deploying, operating, securing, maintaining, troubleshooting, and defending many of the highest-visibility networks in the world. He is a recognized industry leader in the fields of operational security (opsec) and network telemetry, and has an extensive background in security product/feature innovation, devising operational security requirements for network infrastructure devices, and protocol design. His focus is on extending the availability, scalability, and security of the network infrastructure and the applications/services it enables, with an emphasis on flexible and resilient global service delivery capabilities.
youtubeAn Analysis of the DDoS for Bitcoins DDoS Extortion Campaign
pdfYour Bitcoins or Your Site(PDF)
12:00pm - 1:30pmSaint FrancoisNewcomers Lunch - By Invitation OnlySponsors:
12:00pm - 1:30pmHochelaga 1-4Welcome LunchSponsors:
1:30pm - 2:00pm

Le Grand Salon & Marquette

The many uses of NetFlow and flow-like data

Most operators know that flow (NetFlow, IPFIX, and sFlow) data can be collected from routers and switches, and can be used for network cost and planning analysis, attack detection, and peering analysis. But bulk flow data just shows the traffic to-and-from. With 'augmented' flow from load balancers, servers, and sensors listening on taps and span ports, new questions can be answered around performance and security of the infrastructure and application traffic. In this talk, we'll do a brief review of flow and 'classic' flow analytics and use cases; cover some of the source some sources of flow that are not commonly looked at but may be available (including OSS sensor software and configs, load balancers, IDS sensors, server-side monitoring software, and nginx logs); and will discuss a number of real-world use cases enabled with 'augmented flow': - Detecting application-level attacks - Peer analysis by performance, including 'peering' beyond the first hop AS - 'Lightweight' network-viewed Application Performance Monitoring, to help users understand if they are seeing application or network-layer issues - Detecting and exposing customer security issues (or for web companies, compromised servers or end user devices) - E-commerce bot detection All of the tools we'll cover for gathering the augmented flow data (nprobe, and 3 packages being released by Kentik for flow -> top talkers, nginx logs -> flow, and bro logs -> flow) are free and/or open source. [Going light on demo and configurations, this is a 30-45 minute talk, or with more configurations and demos of some of the functionality, it would be 45-60 minutes.]

View full abstract page.
Speakers:

  • Avi Freedman, Kentik
  • Mr. Avi Freedman founded Kentik, Inc. and serves as its Chief Executive Officer. Mr. Freedman serves as Technical Advisor of ServerCentral, Inc. Mr. Freedman served as the Chief Technology Officer of ServerCentral, Inc. Mr. Freedman had a number of roles at Akamai Technologies from October 1999 to August 2009, including Vice President of Network Architecture and Infrastructure, and Chief Network Scientist. Mr. Freedman was most recently Distinguished Engineer and Chief Network Scientist for Akamai, where he oversaw the development and evolution of Akamai's network and content delivery technologies to new applications. He led the growth and management of Akamai's globally distributed network from 250 networks to more than 1000 networks. Mr. Freedman served as Vice President of Engineering at AboveNet, and in 1992, he founded Netaxs, the first ISP in Philadelphia.
pdfThe many uses of NetFlow and flow-like data(PDF)
youtubeThe many uses of NetFlow and flow-like data
2:00pm - 2:45pm

Le Grand Salon & Marquette

NANOG Board Candidates

Introductions of the 2015 NANOG Board candidates with a short presentation by each, followed by Q&A, time permitting.

View full abstract page.
Moderators:
  • William Charnock, PacketFabric.
pdfNANOG Board Candidates(PDF)
youtubeNANOG Board Candidates
2:30pm - 3:15pmConvention FoyerBreakSponsors:
3:00pm - 5:00pm

Le Grand Salon & Marquette

Data Center Track

The Data Center Track at NANOG 65 will build on a number of important data center themes. We will open with a presentation by Josh Rabina, Co-President of Sentinel Data Centers, entitled “Overbuilt & Overbought?: Current Trends in Data Center Provisioning”, who will explore the concepts of low vs. high resiliency deployments and stranded density. The main points of the presentation will include: 1. Low resiliency deployments - Exploring options for combined high and low density deployments and how low resiliency specification can lead to wildly variant performance levels. The data center industry should begin to focus more acutely on “mean time to repair” (MTTR) on a component by component level and adjust low resiliency designs accordingly. Examples will be provided of different reference designs, all “N”, but with very different MTTR implications. 2. Stranded density - How overprovisioning has been among the single largest cost factors impacting the industry (likely greater than the individual impact of rental pricing, power cost, PUE, etc.). How does a provider avoid over provisioning, while maintaining flexibility and managing existing stranded density? The presentation will be followed by a panel discussion around the presented topics exploring the advantages and disadvantages to customers. Participants on this panel will include Todd Schreiber, Director of Data Center Strategy, Architecture, and Software at Bloomberg LP, Jason van Gaal, CEO of ROOT Data Center, and Don Beaty, Founder of DLB Associates. Finally, we will have a series of four (4) 15 minute lightning talks as follows: Michel Chartier, Kelvin-Emtech - “Minimizing Environmental and Geo Risks in Site Selection” Jason van Gaal, ROOT Data Center - “Achieving Data Center Efficiency: A Practical Approach” Marc Pare, CloudOps - “Current State of Cloud in Canada” Gabe Cole, RTE Group - “Open-IX Data Center Standards - Future Initiatives”

View full abstract page.
Moderators:

  • Gabe Cole, RTE Group, Inc.
  • Gabe Cole Chair, Data Center Standards Open-IX Gabe Cole, founder of RTE Group, is an accomplished technology executive with over 20 years’ experience across data center, network, and cloud. Gabe assists service providers and enterprises with converged network and data center design, procurement, and migration including colocation and the XaaS stack of services. He has personally worked on over 150 engagements in North America and EMEA. He is an expert at balancing reliability and performance requirements against technology and business limitations to consistently increase performance at progressively lower price points. In addition, he is frequently called on by investors for assistance with data centers, telco hotels, data center technologies, and services companies.
pdfData Center Track(PDF)
youtubeData Center Track
3:00pm - 5:00pm

Jolliet

DNS Track

DNS-OARC intro Keith Mitchell 10 Minutes OpenDNS; Managing DDoS Attacks Brian Somers 30 Minutes F-root Anycast Research using RIPE Atlas Ray Bellis 30 Minutes Impact of unknown EDNS options on the DNS Eddie Winstead 10 Minutes Benchmarking of authoritative DNS servers and DNSSEC impact assessment Tomas Hlavacek 30 Minutes Root KSK Rollover Geoff Huston 10 Minutes

View full abstract page.
Speakers:
  • Ray Bellis, Internet Systems Consortium.
  • Tomas Hlavacek, CZ.NIC.
  • Geoff Huston, APNIC.
  • Keith Mithcell, DNS-OARC.
  • Brian Somers, OpenDNS .
  • Duane Wessels, VeriSign
  • Duane Wessels became interested in web caching in 1994 as a topic for his master's thesis in telecommunications at the University of Colorado, Boulder. He worked with members of the Harvest research project to develop web caching software. After the departure of other members to industry jobs, he continued the software development under the name Squid. Another significant part of Duane's research with the National Laboratory for Applied Network Research has been the operation of 6 to 8 large caches throughout the U.S. These caches receive requests from hundreds of other caches, all connected in a "global cache mesh."
  • Eddie Winstead.
pdfDNS Track(PDF)
pdfDNS Track(PDF)
pdfDNS Track(PDF)
pdfDNS Track(PDF)
pdfDNS Track(PDF)
pdfDNS Track(PDF)
youtubeDNS Track
5:15pm - 6:45pmHochelaga 1-4Peering PersonalsSponsors:
7:00pm - 11:00pmOffsite - Cirque EloizeSocial
Additional information
Sponsors:
Tuesday, October 6 2015
Time/Webcast:Room:Topic/Abstract:Presenter/Sponsor:Presentation Files:
8:30am - 9:30amSaint FrancoisMembers Breakfast - By Invitation Only
8:30am - 9:30amHochelaga 1-4Power BreakfastSponsors:
8:30am - 5:00pmConvention FloorRegistration
9:30am - 11:00amDuluth & Mackenzie

DDoS Tutorial

This talk covers the principles and particular implementations of DDoS. It goes in detail as to what are the bottlenecks that are generally exploited/overloaded, the attack types and the solutions to those.

View full abstract page.
Speakers:

  • Krassimir Tzvetanov, A10 Networks, Inc.
  • Over the past 7 years I worked on abuse/DDoS mitigation Yahoo! and Cisco. At Cisco I also provided consulting for a number of customers on security issues (including DDoS and threat intelligence related to it). Currently at A10 I focus on threat intelligence and product development for security products, as well as serve as high tier escalation point for high profile customers. I have presented at a number of conference on security and DDoS topics. I have ran or co-ran 3 Security Tracks at NANOG so far and I would like to see it have more security content, which I can both arrange for and deliver. I have desire to build a full day security focused track and I can recruit the presenters and deliver the content. More details about my work history can be found on my LinkedIn profile.
pdfDDoS Tutorial(PDF)
9:30am - 11:00am

Jolliet

Demystifying Pros & Cons of large Scale BGP RR deployments

In modern ISP deployments, the number of BGP speakers inside a ISP's core can be more than one thousand and can spread across different geographical locations . To scale with the increasing number of routers, large ISPs have developed and used more scalable i-BGP architectures such as BGP route reflectors (RR) without thorough analysis of their design considerations. This lack of such analysis escalated interests and concerns on BGP performance inside a large ISP; BGP performance inside large ISPs is no longer simple to understand and can potentially have a noticeable impact on the end-to-end data plane performance & customer service impacts. Some positives of RR implementations include reduced networking provisioning cost, reduced memory usage for storing routing tables, and reduced number of update messages generated inside an ISP. However, these benefits come at some serious cost; there are also negative side effects on both routing correctness and routing performance due to RR deployments. The objective of this Tutorial is to provide a thorough understanding of benefits & challenges of large scale RR deployments. It will try to address the concerns from the Internet community regarding the i-BGP topological complexities. The tutorial will also explain the BGP optimization techniques, BGP designs & new features available from different vendors to improve or negate the challenges faced in RR deployments.

View full abstract page.
Speakers:

  • rohit bothra, Brocade Communications
  • Rohit Bothra is a Staff Engineer at Vyatta - a Brocade company, specializing in virtual network appliances, system & solution testing, performance & scale measurement of virtual systems. He has over nine years of rich experience in Networking industry.He has worked with many major service providers in the Asia Pacific region in the area of IP, IPv6, MPLS. His area of expertise includes NFV, VNFs, Routing Protocols, MPLS, IPv6, HA Systems, Routing Platforms, Network Security, Network Operations, providing Network deployment solutions to different service provider customers. In the past, he has represented Cisco at APNIC conferences. He is passionate about learning new technologies.
pdfDemystifying Pros & Cons of large Scale BGP RR deployments(PDF)
youtubeDemystifying Pros & Cons of large Scale BGP RR deployments
9:30am - 11:00am

Le Grand Salon & Marquette

NetDevOps - Ansible 101 to network nirvana

Tired of pasting repetitive configuration blocks from notepad? Fearful that hipster 20-somethings will automate your career? Wondering if it's time to hunker down and really learn a scripting language? Automation, configuration management, unit testing, continuous integration ... all of these methodologies are nothing new and over the past decade, these traditional software development practices have been co-opted into the operations world. While these approaches continue to be all the rage for compute infrastructure admins, what about networking professionals? This tutorial will start with a brief introduction to key DevOps concepts along with how these can be applied to networks. The bulk of the presented material will focus on a beginners guide to using Ansible, a well known open source automation tool. True to NANOG guidelines, examples shown will be vendor agnostic and geared to be applicable to common environments. With time availability, advanced examples which connect into an IPAM system or multi-vendor protocol testing will be quickly reviewed. The end goal of this tutorial is to wrap with comfortable working knowledge of basic Ansible usage and how you might apply it - it is only the tip of the iceberg in networking automation, but crucial to getting started. Before the session, please take time to follow instructions and install needed components at: https://github.com/bronwynlewis/nanog65-automation-tutorial

View full abstract page.
Speakers:

  • Bronwyn Lewis, Packet Clearing House
  • Bronwyn Lewis is an engineer at non-profit Packet Clearing House. Part network engineer and part sysadmin, her work includes provisioning and deployment of DNS nodes around the world, as well as general IT and automation. Prior to PCH, Bronwyn managed operations & various technical projects at a market research start-up in Los Angeles. She studied international rights, justice, and governance issues at the New School in NYC.

  • Matt Peterson, Cumulus Networks
  • Matt Peterson works within the office of the CTO at Cumulus Networks. At Cumulus, he built the initial customer experience team, and is responsible for technical evangelism. Prior to Cumulus, Matt ran network & datacenter operations at Tumblr & Square. He also founded the first non-profit Internet Exchange within San Francisco, known as SFMIX. His work has been presented at numerous industry events, including APRICOT, BSDcon, Defcon, & CCC Camp.
youtubeNetDevOps - Ansible 101 to network nirvana
11:00am - 11:30amConvention FoyerBreakSponsors:
11:30am - 1:00pmJolliet

Security Track

* BGPuma: Are You Being Route Hijacked? Leigh Metcalf * Monitor and alert on filtered prefixes Tim Evens * New IRR analysis tools for operators Job Snijders * Using BGPstream for route hijack detection Andree Toonk * BGPuma Demo Leigh Metcalf * Panel: "How can we work together to improve security and resilience of the global routing system?" Moderator: Andrei Robachevsky Panelists: - Job Snijders (NTT) - Rob Hagens (Zayo) - Andre Toonk

View full abstract page.
Speakers:

  • Krassimir Tzvetanov, A10 Networks, Inc.
  • Over the past 7 years I worked on abuse/DDoS mitigation Yahoo! and Cisco. At Cisco I also provided consulting for a number of customers on security issues (including DDoS and threat intelligence related to it). Currently at A10 I focus on threat intelligence and product development for security products, as well as serve as high tier escalation point for high profile customers. I have presented at a number of conference on security and DDoS topics. I have ran or co-ran 3 Security Tracks at NANOG so far and I would like to see it have more security content, which I can both arrange for and deliver. I have desire to build a full day security focused track and I can recruit the presenters and deliver the content. More details about my work history can be found on my LinkedIn profile.
pdfSecurity Track(PDF)
11:30am - 1:00pm

Le Grand Salon & Marquette

Wi-Fi: Fundamentals, Design and Troubleshooting

Wireless design is often over-looked when rolling out or upgrading Wi-Fi networks. Ignoring the fundamentals of RF and Wi-Fi behaviour during the design phase or not even bothering with a design results in user frustration and unnecessary troubleshooting for operations teams. During this tutorial, you will: 1) review the fundamentals of RF, Wi-Fi terminology and concepts; 2) learn best practices for designing Wi-Fi networks; and 3) learn how to troubleshoot Wi-Fi networks using various tools including deciphering packet captures. After attending this tutorial, you will have gained insights that you can apply to your existing networks or upcoming designs.

View full abstract page.
Speakers:

  • Troy Martin, Aerohive Networks
  • Troy Martin is a Systems Engineer at Aerohive Networks, specializing in mobile and cloud technology.  In addition to teaching others how Wi-Fi works, his work includes designing, provisioning, and troubleshooting Wi-Fi networks .  Prior to Aerohive, Troy provided Wi-Fi consulting in multiple verticals including service provider, pharmaceutical, education  and industrial environments such as mines, refineries and chemical plants.  He is passionate about connecting things without wires.  Troy is a P.Eng. who holds Bachelor degrees in both electrical engineering and computer science from the University of Saskatchewan.
pdfWi-Fi: Fundamentals, Design and Troubleshooting(PDF)
youtubeWi-Fi: Fundamentals, Design and Troubleshooting
1:00pm - 2:00pmHochelaga 1-4LunchSponsors:
2:30pm - 3:15pm

Le Grand Salon & Marquette

10 Weird (and Possibly Useful) Things You Didn't Know about International Networks

Web advertisers know that people can't resist lists. Online lists (e.g., Ten Celebrities You Won't Believe had Plastic Surgery) have suckered in many a viewer. Structured to be fast-paced and fun like those web lists, this presentation will address questions bandwidth buyers frequently ask about carriers. Think online clickbait, but with less Angelina Jolie and more scatterplot charts. The presentation will share primary data recently gathered by TeleGeography on IPT and capacity prices. Issues to be addressed include: What are carrier pricing strategies for 100G circuits, and how do 100G prices compare with 10G? Where have IPT price declines stalled? Where will capacity prices soon tank? Does anyone buy IRUs anymore? The presentation will also address how network builds could affect the business plans of NANOG attendees. Items to be covered include: How fast are content providers buying up capacity, and what are the spillover effects for the rest of the industry? What parts of the world are causing the most headaches for companies expanding their networks? And, finally: just how much of a threat do sharks pose to our critical undersea cable infrastructure?

View full abstract page.
Speakers:

  • Tim Stronge, TeleGeography
  • Tim Stronge is Vice President of Research at TeleGeography. His areas of expertise include international voice traffic, terrestrial and submarine cable systems, and international bandwidth markets. Since joining TeleGeography in 1996, Tim has served as a principal analyst in most areas of research, including network infrastructure, bandwidth demand modeling, cross-border traffic flows, and telecom services pricing. He holds a Master's degree in International Economics from John Hopkins University and a B.A. from the College of William and Mary.
youtube10 Weird Things You Didn't Know about International Networks
3:15pm - 4:00pm

Le Grand Salon & Marquette

Monitor BGP using open source OpenBMP and Apache Kafka

BGP and OSPF/IS-IS monitoring is commonly implemented using VTY scripting of CLI output or by using passive BGP peering sessions. Scripting CLI output can be slow and cumbersome in terms of credential management and it can be prohibitive when access is restricted. In addition, CLI output is specific to the router OS resulting in the scripter having to constantly maintain various router implementations. BGP peering is complex to implement and overkill for most use-cases when only monitoring of prefixes or viewing of RIB tables are needed. Existing tools can be difficult to deploy and have complex API's when interacting with both RAW and parsed BGP data. We will show that OpenBMP breaks down the traditional barrier of complexity in interacting with BGP data. We will discuss why Kafka is being used and review features that Kafka provides over other existing methods to convey real-time parsed and raw BMP/BGP data. We will cover use-cases and examples of different types of OpenBMP/Kafka consumers. We will show how a network engineer can interact with real-time live BGP data using simple BASH and AWK scripting or by using CQL/SQL queries.

View full abstract page.
Speakers:

  • Tim Evens, Cisco
  • Tim Evens has over 19 years of experience as a network engineer and programmer in wide range of industries including carrier and internet service providers, financial trading, healthcare, retail and technology. Tim has been active as a double CCIE for 14 years. In his current position, Tim works in the Chief Technology and Architecture Office at Cisco and is responsible for establishing inter-networking data analytics. He is the maintainer of OpenBMP (www.openbmp.org).
pdfMonitor BGP using open source OpenBMP and Apache Kafka(PDF)
youtubeMonitor BGP using open source OpenBMP and Apache Kafka
4:00pm - 4:30pmConvention FoyerBreakSponsors:
4:30pm - 5:00pm

Le Grand Salon & Marquette

Building a smallish DC...for the rest of us

We're not Facebook, Amazon, Google or Microsoft, but we still need to build new Data Centers. They may not be huge, but we still want to do it right, sized for today and able to easily scale. It had to be vendor agnostic, use standard protocols, accommodate legacy systems, improve security, reduce costs, blend seamlessly into cloud, only rock the boat within Engineering so far, integrate into legacy sites and not cause my cell to ring at 2am. In other words, the usual impossible asks. What sucked about our legacy L2/L3 design, the initial MPLS VPN design that didn't work and the current L3 design. Here is a story of how we rebuilt our Data Centers, balancing costs, legacy systems, developers/users, conflicting priorities, public clouds and inheriting a project already in progress. In the end, the configs were the easy part.

View full abstract page.
Speakers:

  • Karl Brumund, Dyn
  • Karl Brumund has been involved in networking and telecommunications for over 25 years. He is presently a Principal Network Engineer and Technical Owner of Network Engineering with Dyn, upgrading and expanding Dyn's global Internet footprint. Prior to Dyn, he held a variety of positions with Rogers Communications, Futureway, Nortel, UUNET and IBM.
pdfBuilding a smallish DC...for the rest of us(PDF)
youtubeBuilding a smallish DC...for the rest of us
5:00pm - 5:30pm

Le Grand Salon & Marquette

Running MPLS efficiently in ring networks

MPLS is widely deployed for the key benefits it offers: traffic engineering and fast reroute. MPLS signaling protocols have been built in a very general-purpose manner. Such generality renders them somewhat inefficient in deployment for special network topologies like a ring. MPLS deployments in ring topologies, that are common in access / aggregation networks, end up being more complicated to configure and manage than is strictly necessary. The scarce bandwidth also ends up being used inefficiently. This talk shall: - Describe why MPLS ring topologies are special and thus merit tailor-made MPLS extensions - Describe the mechanics of how to run MPLS more efficiently in a ring network. The efficiency benefits described would cover both the operational (configuration / management) aspect and also the network utilization aspect. - Illustrate the benefits of the MPLS-for-rings extensions compared to running general-purpose MPLS in a ring.

View full abstract page.
Speakers:

  • Ravi Singh, Juniper Networks
  • Ravi Singh is a principal software engineer in the routing/MPLS team at Juniper Networks. He has implemented various unicast/multicast and MPLS-signaling-protocols functionality. He helped ship the PTX series of Juniper platforms that are getting deployed in the super-core of various tier-1 networks.
pdfRunning MPLS efficiently in ring networks(PDF)
youtubeRunning MPLS efficiently in ring networks
5:30pm - 6:00pm

Le Grand Salon & Marquette

BGPuma -- Border Gateway Protocol Update Metric Analysis

Analyzing BGP update files is an arduous process, especially when one is looking for updates that can affect a given CIDR block. BGPuma was created to search through these files in a reasonable time. Given a list of CIDR blocks, it looks for routes announcing CIDR blocks which match the list, are contained in the list, or are contained by the list. For Autonomous Systems, it looks for routes that are announced by the given system. It combines the speed of SiLK (http://tools.netsa.silk.org/silk/index.html) and the power of the RIPE bgpdump library (https://bitbucket.org/ripencc/bgpdump/wiki/Home) to accomplish this.

View full abstract page.
Speakers:

  • Leigh Metcalf, CERT
  • Leigh Metcalf holds a PhD in Mathematics from Auburn University at Auburn, Alabama. She also has over 10 years of industry experience as a systems engineer, network engineer, and software developer. She is currently a senior member of the technical staff at CERT.
pdfBGPuma -- Border Gateway Protocol Update Metric Analysis(PDF)
youtubeBGPuma -- Border Gateway Protocol Update Metric Analysis
6:00pm - 8:00pmHochelaga 1-4Beer 'n GearSponsors:
8:00pm - 11:00pmOffsite - Bar commerceSocial
Additional information
Sponsors:
Wednesday, October 7 2015
Time/Webcast:Room:Topic/Abstract:Presenter/Sponsor:Presentation Files:
8:30am - 9:30amHochelaga 1-4Power BreakfastSponsors:
8:30am - 6:00pmConvention FloorRegistration
9:30am - 10:00am

Le Grand Salon & Marquette

Whack-a-Mole Routing: The effects of ISP Traffic Engineering on Non-ISP Networks

This will be a discussion about the impact of ISP traffic-engineering on non-ISP networks (CDNs, ADNs, etc). There will be an overview of scenarios that led to inefficient routing and excessive delay. These delays in turn impact End-user Page Time (EPT) and Time To First Byte (TTFB). What is being done by non-ISP networks to address those issues. This includes independent actions by non-ISPs as well as joint ones with their upstream provider(s) or the ISP who made the change. The joint actions will include how to address issues when an ISP that neither the non-ISP nor end customer have a formal relationship with makes a TE or Peering change.

View full abstract page.
Speakers:

  • Rick Casarez, Salesforce
  • Rick Casarez is a Principal Network Engineer in the Applied Research team at Salesforce. Prior to Salesforce he has previously worked at ISPs like UUNet (701) and AboveNet (6461).
pdfWhack-a-Mole Routing(PDF)
youtubeWhack-a-Mole Routing: The effects of ISP Traffic Engineering
10:00am - 10:30am

Le Grand Salon & Marquette

Distributed Route Aggregation on the Global Network (DRAGON)

The Internet routing system faces serious scalability challenges due to the growing number of IP prefixes that needs to be propagated throughout the network. Although IP prefixes are assigned hierarchically and roughly align with geographic regions, today's Border Gateway Protocol (BGP) and operational practices do not exploit opportunities to aggregate routing information. In the talk, I will present DRAGON, a distributed route-aggregation technique whereby nodes analyze BGP routes across different prefixes to determine which of them can be filtered while respecting the routing policies for forwarding data-packets. DRAGON works with BGP, can be deployed incrementally, and offers incentives for Autonomous Systems (ASs) to upgrade their router software. I will illustrate the design of DRAGON through a number of examples and I will present results on its performance. Experiments with realistic AS-level topologies, assignments of IP prefixes, and routing policies show that DRAGON reduces the amount of forwarding and routing state in the ASs by up to 80% with minimal stretch in the lengths of AS-paths traversed by data-packets.

View full abstract page.
Speakers:

  • João Luís Sobrinho, Instituto de Telecomunicações, University of Lisbon
  • João Luís Sobrinho received the Licenciatura and Ph.D. degrees in Electrical and Computer Engineering from Instituto Superior Técnico, Universidade Técnica de Lisboa, Portugal, in 1990 and 1995, respectively. From 1995 through 1997 he was a Member of the Technical Staff at Bell Labs, Lucent Technologies, The Netherlands, working in the area of wireless communications. Since 1997, he has been with the Department of Electrical and Computer Engineering, Instituto Superior Técnico, Universidade de Lisboa, where he is currently an Associate Professor, and with the Instituto de Telecomunicações, where he is currently a senior Researcher. His present research interests cover distributed algorithms in networking, with an emphasis on routing. He won a Best PhD Student Paper Award at PIMRC 1994, the 2006 IEEE Communications Society William R. Bennett Prize, and an 2015 IRTF Applied Networking Research Prize. He is a senior member of the IEEE and a member of the ACM.
pdfDistributed Route Aggregation(PDF)
youtubeDistributed Route Aggregation on the Global Network (DRAGON)
10:30am - 11:00am

Le Grand Salon & Marquette

Cloudy with a chance of Breach: Forecasting Cyber Security Incidents

[This work will first appear at the USENIX Security Symposium in August] In this presentation we characterize the extent to which cyber security incidents, such as those referenced by Verizon in its annual Data Breach Investigations Reports (DBIR), can be predicted based on externally observable properties of an organization’s network. We seek to proactively forecast an organization’s likelihood of a security incident. To accomplish this goal, we collect 258 externally measurable features about an organization’s network from two main categories: mismanagement symptoms, such as misconfigured DNS or BGP within a network, and malicious activity time series, which include spam, phishing, and scanning activity sourced from these organizations. Using these features we train and test a Random Forest (RF) classifier against more than 1,000 incident reports taken from the VERIS community database, Hackmageddon, and the Web Hacking Incidents Database that occurred between mid-2013 and the end of 2014. The resulting classifier is able to achieve a 90% True Positive (TP) rate, a 10% False Positive (FP) rate, and an overall 90% accuracy.

View full abstract page.
Speakers:

  • Manish Karir, QuadMetrics
  • Manish Karir has been an active participant in the NANOG community since 2005. He is a frequent presenter at NANOG and over the past 10 years has had the opportunity to present over 11 times on a wide variety of topics including, BGP analysis tools, traffic analysis and visualization tools, darknet traffic research, IRR enhancements, network reputation and RBL analysis, IPv4 address sharing techniques, and cybersecurity posture metrics. His research interests include Internet measurement studies, and large scale network data collection and analysis. Manish is currently the Chief Technology Officer at QuadMetrics.
pdfCloudy with a chance of Breach(PDF)
youtubeCloudy with a chance of Breach: Forecasting Cyber Security Incidents
11:00am - 11:30amConvention FoyerBreakSponsors:
11:30am - 12:00pm

Le Grand Salon & Marquette

Optimal routing vs. Route Reflector VNF - reconcile the fire with water

The NFV technology promises cost reduction by use of commoditized compute hardware, flexible CPU and memory resources allocation, and reduced cost. The big part of this advantages come form fact that VNF is instantiated in datacenter. When VNF is used for routing control plane nodes, like BGP RR, topological position of it in network may impact routing decision. Traditionally, RR are located startegically in network in a way to optimize routing and reduce unnecessary transport of traffic flows by own network. The data centers may not be located in same places where RR traditional are. On this talk a solution for overcome above problem and ensure optimal routing decision while benefit from instantiation RR VNF in datacenter of anylocation. This solution is based on traditional BGP-4 implementation of RR and do not requires ORR. Security and DDoS protection aspects of solution are also covered.

View full abstract page.
Speakers:

  • Rafal Szarecki, Juniper Networks
  • Rafal Szarecki (JNCIE 136) has 19 years of experience in networking industry. He has started his career working for ISP in Poland as Network Operation Centre stuff member in 1996. Later in his career he work in Customer Support consultant and then as Network Architect in Ericsson. In 2006 Rafal joined Juniper Networks Professional Services Team and later become Principal Consultant. He has worked on over 50 different projects, with more then 20 Telecoms and ISP including some of global reach such British Telecom, Telefonica, Telenor. He designed and/or helped in deployment over 20 networks in multiple countries in Europe and Middle East. During this time he was also active member of PLNOG. Currently Rafal held position of Solution Architect in Routing Product Team in Juniper Networks in Sunnyvale, CA.
pdfOptimal routing vs. Route Reflector VNF(PDF)
youtubeOptimal routing vs. Route Reflector VNF
12:00pm - 12:30pm

le Grand Salon & Marquette

The future of North American Regional BCOP

The future of North American Regional BCOP. An update on what has been done since the last NANOG. Update on Global Coordination including progress with the IETF process and RFC editors. Regional options pros/cons and community feedback. Chris Grundeman and I will give a joint presentation (if approved). Requesting a 20 min slot. Cheers, Aaron

View full abstract page.
Speakers:
  • Chris Grundemann.
  • Aaron Hughes, 6connect.
pdfThe future of North American Regional BCOP(PDF)
youtubeThe future of North American Regional BCOP
12:30pm - 1:00pm

Le Grand Salon & Marquette

Lightning Talks

The Program Committee will be accepting Lightning Talk submissions for 10-minute speaking slots, via pc.nanog.org, beginning Sunday afternoon, Oct 4. You'll be notified the night before the Lightning Talk session if you've been selected.

View full abstract page.
Speakers:
  • NANOG Lightning Talks, NANOG Community.
pdfLightning Talks(PDF)
pdfLightning Talks(PDF)
pdfLightning Talks(PDF)
youtubeLightning Talks
1:00pm - 2:30pm Lunch (On Your Own)
2:30pm - 4:00pmLe Grand Salon & Marquette

Peering Track

Peering: Boulevard of Broken Networks Speaker: John Nistor, TorIX Panel: FCC Open Internet Order Moderator - Patrick Gilmore, Markley Group Panelists: Hank Hultquist - AT&T, Vice President Federal Regulatory David E Young - Verizon, Exec Director Federal Govt Relations Joseph Cavender - Level 3, Vice President & Assistant General Counsel, Federal Affairs

View full abstract page.
Moderators:

  • Patrick Gilmore, Markley Group
  • Patrick Gilmore is Markley Group’s Chief Technology Officer (CTO), responsible for Markley’s Cloud Computing and Infrastructure as a Service offerings. Markley Group runs New England’s largest and longest operating multi-tenant, mission-critical telecommunications and data center. Prior to Markley, Patrick was Chief Network Architect of Akamai Technologies, where he worked for over 13 years. At Akamai, Patrick’s group was responsible for managing peering and capacity for the world’s largest CDN, one that served approximately 25% of all traffic on the internet. Patrick has a degree in mathematics from UCLA. Patrick is past member of the NANOG board of directors, and helped move NANOG to an independent organization. Patrick is currently on the Board of Directors for the Seattle Internet Exchange and the London Internet Exchange, two of the largest Internet exchanges in the world, as well as the PeeringDB, where he is Chairman.

  • Sylvie LaPerriere, Google Inc.
  • Sylvie LaPerrière is a builder of the Global Internet. An expert in peering and international network interconnection agreements, she is a strategic negotiator for Google’s Content Distribution and Network team. Before joining Google, she created and then expanded Tata Communications’ Internet backbone in more than 25 global markets, had it circle the globe and achieved settlement-free status. Operating in the international telecommunications arena for two decades, Sylvie has a solid experience developing and managing Internet, data and mobile services. As negotiator of business agreements, she has an extensive experience dealing with global, regional and national telecom and Internet operators.
Panelists:

  • Joseph Cavender, Level 3 Communications
  • Joe Cavender is vice president and assistant general counsel for Federal Affairs at Level 3 Communications, LLC, where he leads the U.S. federal public policy team. Before joining Level 3 in March 2013, he worked at the FCC, primarily in the Wireline Competition Bureau’s Telecommunications Access Policy Division. Prior to his government service, Cavender was an associate at Harris, Wiltshire & Grannis LLP in Washington, D.C. Cavender earned a B.A. in Philosophy from Yale University and a J.D. from the University of Chicago. Following law school, he clerked for the Hon. James B. Loken on the United States Court of Appeals for the Eighth Circuit.

  • Hank Hultquist, AT&T
  • Hank joined AT&T (then SBC) in 2004. He represents AT&T at the Federal Communications Commission (FCC) on a number of issues including broadband and Internet policy, video and media policy, intercarrier compensation, and universal service. He is a member of the Board of Directors and of the Wireless Communications Association International (WCAI) and serves on the North American Numbering Council (NANC). Prior to joining AT&T, he was with MCI for eight years. Hank is a graduate of the George Mason University School of Law and the Georgetown University School of Foreign Service.

  • David E. Young, Verizon
  • David E. Young, Vice President, Verizon Public Policy, is responsible for identifying and assessing emerging issues, developing corporate positions on Internet and Technology industry policy matters, and assessing key technology and communications industry trends. He is also responsible for developing relationships with high technology industry members, technology associations, research institutes, and think tanks. Previously, he was responsible for items before the FCC dealing with broadband and emerging issues. Prior to 2000, he spent six years working in Verizon’s Research and Development (R&D) group on many advanced technologies including VoIP, data network architectures, and audio, video and image compression. He has been awarded ten U.S. government patents for his R&D work. David is a member of the IEEE and IEEE Communications Society. David holds a bachelor’s degree in Electrical Engineering from Stevens Institute of Technology.
Speakers:

  • Jon Nistor, TorIX
  • For the past 14 years, Mr. Nistor has been involved in numerous aspects of TorIX, such as day to day operations, design and automation, and serving on the board. He has been involved with TorIX since its ages of sub 1Gbps to over 180Gbps now. Mr. Nistor is also Systems Engineer with Cisco Systems during the day.
pdfPeering Track(PDF)
pdfPeering Track(PDF)
2:30pm - 4:00pm

Duluth & Mackenzie

Strategies of packet buffering inside Routers

In this talk I describe buffering architecture, starting form shared memory used in devices build around single centralized switch (SoC) and then focusing on distributed architectures more common used in today's routers. The OQ, IQ, VoQ and CIOQ buffering strategies are described and explained, following by strengths and limitation of each. Finally impact of chosen architecture on ASIC design would be briefly evaluated. The intention is to provide knowledge about foundation of routers system architecture, that could be useful to troubleshoot certain network events that manifest as packet loss somwhere inside router. The whole talk is based on theoretical models, not related to any specific vendor's product. However each product follows (w/ some deviation) one of discussed model - so knowledge provided would be relevant to operation of this product and understand their internal as well as deviation in behavior of different product when exposed to the same traffic.

View full abstract page.
Speakers:

  • Rafal Szarecki, Juniper Networks
  • Rafal Szarecki (JNCIE 136) has 19 years of experience in networking industry. He has started his career working for ISP in Poland as Network Operation Centre stuff member in 1996. Later in his career he work in Customer Support consultant and then as Network Architect in Ericsson. In 2006 Rafal joined Juniper Networks Professional Services Team and later become Principal Consultant. He has worked on over 50 different projects, with more then 20 Telecoms and ISP including some of global reach such British Telecom, Telefonica, Telenor. He designed and/or helped in deployment over 20 networks in multiple countries in Europe and Middle East. During this time he was also active member of PLNOG. Currently Rafal held position of Solution Architect in Routing Product Team in Juniper Networks in Sunnyvale, CA.
pdfStrategies of packet buffering inside Routers(PDF)
youtubeStrategies of packet buffering inside Routers
4:00pm - 4:30pmHochelaga 3Break
4:30pm - 5:00pm

Le Grand Salon & Marquette

Proactive Network Configuration Validation with Batfish

Batfish is an open-source network configuration analysis tool in active development produced jointly by researchers at University of California, Los Angeles; University of Southern California; and Microsoft Research. Though its individual modules have various applications, its primary purpose is to detect bugs in network configurations. Batfish takes as input a set of network configurations, and an environment, which consists of a set of (in)active links and a set of external BGP advertisements. Users are able to ask customized queries about the control plane using Batfish's domain-specific query language e.g. whether all loopback addresses are being advertised into OSPF, or whether all route policies attached to eBGP neighbors apply a particular community to incoming routes. Batfish also is able to compute the convergent data plane for a network, which provides further query facilities. Given the data plane, users can employ an off-the-shelf data plane checker or use Batfish's data-plane queries to check common properties such as reachability/black holes, loops, etc, as well as novel properties (introduced at NSDI'15) regarding equivalence of multipath routes, fault-tolerance, and unique delegation of customer address space, with more to come.

View full abstract page.
Speakers:

  • Ari Fogel, UCLA
  • I am a 5th-year Ph.D. student at the Department of Computer Science at University of California, Los Angeles. My research has focused on static analysis of network configurations, with an eye toward detecting deviations from operator intent. On this work I have collaborated with researchers at University of Southern California and Microsoft Research, where I am currently interning.
  • ratul mahajan, microsoft research.
  • Todd Millstein, University of California, Los Angeles.
  • Meg Walraed-Sullivan, Space Exploration Technologies.
pdfProactive Network Configuration Validation with Batfish(PDF)
youtubeProactive Network Configuration Validation with Batfish
ptxProactive Network Configuration Validation with Batfish (slides)(PTX)
5:00pm - 5:30pm

Le Grand Salon & Marquette

Deploying IPv6 at Scale as an ISP

As a service provider serving nearly 14 million customer connections in Canada across Wireless and Wireline services, TELUS has a vast set of technologies on multiple networks with many supporting systems enabling these services. Having worked for years on deploying IPv6 we are now introducing IPv6 support for our customers in all major Internet services. Our first IPv6 services were delivered in 2012 to ARIN, NANOG and IETF which all met in Vancouver and relied on the TELUS network for connectivity. Three years later, you might not guess it based on the mere 0.53% of Canadian users enabled as of July 10 (according to Google's IPv6 stats) but our deployment is about to scale, particularly for our Residential High Speed Internet. I wish to share what TELUS has learned in our IPv6 efforts including how to establish an IPv6 program, how to set its scope, how to gain executive attention and how to influence other departments to buy-in to IPv6 as necessary for the scope of the many other projects on-going at any point. In short, I will share the valuable lessons we have learned in our IPv6 program which is now coming to fruition, bringing IPv6 to our subscribers. With feedback from the program committee, I will entertain refining this abstract to serve a felt need of that NANOG audience. I would welcome the opportunity to serve the community and help the many ISPs and Enterprises present to succeed in their IPv6 deployment strategies. Regards, Matthew Wilder, P.Eng.

View full abstract page.
Speakers:

  • Clinton Work, TELUS
  • Sr Design Specialist with TELUS since 2001 and based out of Calgary, AB. My focus at TELUS is on MPLS/IP core network support and engineering. I was involved in building the TELUS MPLS core back in 2003 and all the services that we have layered on top since. I have extensive experience with Internet peering, MPLS, ISIS/BGP, L3VPNs, VOIP, and LTE/HSPA+. More recently I have focused on CDN cache server deployments and the TELUS IPV6 program for consumer services (DSL/GPON).
pdfDeploying IPv6 at Scale as an ISP(PDF)
youtubeDeploying IPv6 at Scale as an ISP
5:30pm - 6:00pm

Le Grand Salon & Marquette

Closing Remarks

The 2015 Election Results announced, followed by NANOG 65 Closing Remarks.

View full abstract page.
Speakers:

  • Betty Burke, NANOG Executive Director
  • Currently serving as the NANOG Executive Director, responsible for all aspects of NANOG, reporting to the Board of Directors. Previous 37 years of experience serving in technology, business, and management within the Michigan Information Technology Services, University of Michigan, and Merit Network. Proven leadership and experience in development of strategic and operational plans, creation and implementation of marketing campaign for conference center and high tech facilities including a data center, conference and office building, library and campus fiber assets. Proven operational success through project management, along with leadership through community and team building. - See more at: https://www.nanog.org/governance/bod#sthash.GFpR1G2i.dpuf
youtubeClosing Remarks
6:00pm - 7:00pmHochelaga 1Social
Additional information

 

^ Back to Top