kentik netflow SaaS

www.kentik.com

kentik is a netflow saas that has some serious clue on board

one founder was a founder of Akamai. CTO built netflix CDN. Staff person I talked to used to work at Arbor, etc.

They built their own dbase modeled on Google dremel. 

They are working closely with Luca Deri and ntop. There is flag to export ntop direct to kentik cloud as part of their demo

They integrate with ddos tools including A10. Their netflow system can automatically detect and alert on ddos and then mitigate via your ddos system

They can provide an on-site server for sites with issues about  moving netflow data offsite. They have https access to their cloud servers

Contacts:
Michael Jacobs, Strategic Accounts,  San Francisco, 408.515.9408
Larry Austin, Strategic Account ext, SF, 408.796.1292

Experiences with network automation at Dyn

At NANOG 63 I talked about Kipper, our network automation project at Dyn, and how it aims to align our network configuration lifecycle with the existing continuous integration model used for servers. Since then we have significantly expanded its coverage, added new features and incorporated other teams in our workflow. In this presentation I will describe the current setup and then talk about our challenges, successes and some of the lessons we have learned along the way.

Post IPv4 Depletion Trends

Hijackers are focusing on networks without an ARIN POC - affects legacy/academic networks according to the presenter:
https://www.nanog.org/sites/default/files/Nobile_Post_Ipv4_Depletion.pdf

Automating Maintenance Notifications

Very interesting system for standardizing and parsing maintenance events and getting the information out to the right people, on the right calendars, etc.

Despite all the recent progress around network automation, there's one aspect of our operations that for many remains stuck in a manual past. Most of us deal with maintenance notifications - both those we get and those we send - by throwing people at the problem.

Thousand Eyes

Was told that Thousand Eyes is opening an Austin office with marketing and possibly some training support.

Need to investigate as a possibility for local training and education on TE, especially for other ITS/IT groups who need to understand how to monitor cloud services.

Contact:
Steve Brown, Senior Solutions Engineer, San Francisco, sbrown@thousandeyes.com, 770.335.0354

Utilizing Kea hook points for modern IP addressing (DHCP)

Kea is a new high performance, open source project for DHCP IPv4 and IPv6 addressing.

"No restart required unless physical interfaces change"

Facebook Kea DHCP servers all answer on one IP address
Facebook proposed using Ke dbase to also store Kea config to provide hitless config with no restart

Measurement based inter-domain traffic engineering

This presentation concentrates on an inter-domain traffic engineering scenario for multi-homed stub ASes.

Suffering Withdrawal; an automated approach to connectivity evaluation

Today’s routers generally make themselves more- or less-attractive to transit traffic through operator’s manipulation of their interfaces IGP metrics or overload status. This all-or-nothing method lacks granularity and does not take advantage of the wealth of connectivity and health-check information readily available at the router.

TCP over IP Anycast - Pipe dream or Reality?

The talk will focus on how to route our end users to the closest location serving content -- i.e. to the closest PoP. Traditionally LinkedIn used geo-location based load balancing (with help of DNS) but there are challenging areas with this approach that lead to bad performance for the end user and operational challenges for the LinkedIn site teams.

Blackholing at IXPs: On the Effectiveness of DDoS Mitigation in the Wild

DDoS attacks remain a serious threat not only to the edge of the
Internet but also to the core peering links at Internet Exchange Points.

DNS-based censorship: theory and measurements

As explained in RFC 7754, "Technical Considerations for Internet Service Blocking and Filtering", it is tempting for a censor to attack, not the direct traffic or servers, but the rendezvous systems, the most obvious one being the DNS.

Peering Security and Resiliency

In this presentation we'll talk about BGP peering security and resiliency challenges. First we'll show real-world peering observations from the perspective of a peering router at an IXP. Then we'll give an operational perspective on peering configuration challenges, with a focus on scale an automation.

Everyday practical BGP filtering

Robust BGP filtering is a challenge in and of itself. In this talk
NTT offers unprecedented insight into how today's AS2914
filter-sausage is made.

Avoiding Nation-State Surveillance

When Internet traffic enters a country, it becomes subject to those countries’ laws. As an increasing number of countries pass laws that facilitate mass surveillance, Internet users have more need than ever to determine---and control---which countries their traffic is traversing. To this end, we first conduct a large-scale measurement study to demonstrate that Internet paths often transit countries where laws may make users more vulnerable to surveillance than they would be in their home country.

DNSSEC Tutorial

Eddie Winstead from ISC -- tutorial on DNSSEC.

Root has been signed as of July 2010 ("Layer 9 issue") so only need one trust anchor in your configuration

As of 2016 DNSSEC deployment is much easier. 

Tutorial Everything You Always Wanted to Know About Optical Networking

Richard Steenberger preso on current state of the art for optical networking.

Useful set of slides: https://www.nanog.org/sites/default/files/Steenbergen.Everything_You_Need.pdf

Elliptic curves to the rescue: tackling availability and attack potential in DNSSEC

Notes: Not quite ready for prime time, but moving to better security in the next year or so is on the roadmap. Something to consider if/when making a DNSSEC deployment.

NANOG 67 - Chicago, June 13-15, 2016
894 attendees

Network Support for TCP Fast Open

Notes: Biggest issue was firewalls incorrectly dropping TCP packets with these options. Broken code in firewalls in the systems that they investigated. So middleboxes are making improvements difficult. APprox 20 percent of sites had issues with middlebox-induced failures.