NZNOG 2018 Conference Speakers

Talks currently accepted in to conference programme (Thursday and Friday)


 Company

 Speaker(s)

 Talk title

Facebook

Jose Leitao & Daniel Rodriguez

Build your own version of NetNorad using OpenSource tools

Minister of Communications
Clare Curran
Conference opening

Netflix

Ian Chiu

How Netflix works (if you own and operate a network in New Zealand)

EOLO

 Francesco Alberti

The "BLU" Project: Past, present and future

The talk presents "BLU", an SDN framework developed by Eolo SpA, an Italian Fixed Wireless Access ISP. The wireless nature of the majority of backhaul links, the rather frequent topology changes and a high mesh factor are some of the specific features of the Eolo network. BLU aims at efficiently solving the challenges imposed by unconventional traits of our network. We will discuss the solutions we adopted so far, ranging from hardware to protocols and algorithms, and show how we manage to adapt and combine widely exploited frameworks within our application domain.


NZIX

Joe Wooller and Tim Raphael

NZIX Update and Metrics

Hawaiki

Hamid Maani

Hawaiki Submarine cable

Vodafone / TGA

Phil Hunt

TGA Submarine cable

OpenNMS

Tarus Balog

OpenNMS monitoring platform


OpenNMS is the world's first enterprise and carrier-grade network monitoring platform built using the open-source development model. It was founded in 1999 with the first code publicly released in March of 2000. Since then it has grown to become the choice for monitoring when scale is an issue, and its open source nature means that it can be easily extended. While many commercial solutions require companies to change their processes to fit the application, OpenNMS can be modified to fit those processes. As it is how a company deals with its customers that make it unique, this can create a market advantage.


This presentation will serve as a quick overview of the OpenNMS platform. It will include a discussion of the four main functional areas:


1) Events and Notification Management: OpenNMS can gather events from a number of sources, process and reduce them, as well as generating notifications. There are systems in production that can handle 10,000 events per second.


2) Service Assurance: OpenNMS was started in an era when people were very concerned with service level agreements (SLAs). OpenNMS can perform synthetic transactions to make sure network services, from passing traffic to web-based applicatiosn, are operating properly and identify root causes when they are not.


3) Performance Data Collection: As the Internet of Things becomes a reality, the amount of data available about the performance of the network will grow exponentially. OpenNMS supports the time series database Newts (https://newts.io) which is built on Apache Cassandra, and can scale to billions of metrics.


4) Provisioning: One thing that a lot of applications leave out is a way to actually provision the monitoring system. When you are dealing with hundreds of thousands of devices you have to have a way to automate what it being monitoring and have the ability to audit it to insure nothing gets missed. OpenNMS has a strong provisioning system to address this issue.

Arista Networks

Richard Bayliss

Scaling Beyond 100G with 400G and 800G Ethernet

Bandwidth requirements continue to grow at an exponential pace driven by UHD Content, IOT, Serverless Compute, Machine Learning (ML) and Artificial Intelligence (AI). Both Service Providers and new Cloud operators are racing to adopt the latest technology to keep up with the bandwidth demand in a cost effective manner. With network bandwidth requirements doubling every two years, SPs and Cloud Titans will need 400GE/800GE to help them scale the cloud DC to remain competitive before 2019. 

In this vendor neutral session, we will highlight new 400GE/800GE technology, particularly OSFP and QSFP-DD, and analyse network architecture evolution options. We will also discuss the new 25G and 50G standards and demonstrate how they are the current most economic way to address bandwidth scaling today with the added benefit of providing a future proof pathway to 400G Ethernet and beyond.

Rich is the Director for Systems Engineering at Arista Networks in Asia-Pacific, where he leads efforts to migrate complex networks to simplified, software-driven and agile cloud networks. With businesses keen to capture the inherent advantages of cloud, Rich places particular emphasis on the necessary changes to achieve agile operations and web-scale architectures. 

Rich is internationally recognised for his contribution to the development of new networking standards and architectures. Prior to Arista, Rich held senior positions with Juniper Networks, including CTO for APAC SP and worldwide Chief Architect for Edge Routing, as well as operational and engineering roles at Australian and global service providers.

Vocus

Egor Krivosheev

Network automation

Network automation has been a hot topic over the few years in the industry. With the emergence of the tools like Ansible and Salt Stack, it has become much easier to automate routine operations.

We are at Vocus Networks team, and perhaps many others find it challenging to implement network automation at scale in a large multi-vendor environment. During my talk, I will take you through our journey trying to highlight new developments and best practices such as:

● Inventory and Autodiscovery

● APIs

● Programming languages and libraries

● NETCONF / YANG / Openconfig /gRPC

I will be providing real-world usage examples for technologies mentioned above without going into many technical details.

Kraig Winters

Red Cross

NZ Red Cross IT and Telecommunications Emergency Response Unit

Cisco

Rada Stanic

Model Driven Telemetry – Foundation for Big Data Analytics


Abstract:


With the increasing needs to gain full network visibility at scale, Model Driven Telemetry (MDT) is rising to this challenge. MDT is helping transform the landscape of network monitoring into a Big Data playground. Today's largest Web providers are already reaping the benefits of telemetry and the trend is equally gaining momentum with service providers globally.


This session will cover plethora of exciting open source and new technologies (YANG data models, Kafka, Prometheus, GPB etc.) that allow SP’s operations teams reap the benefits of streaming telemetry. Even if it is as simple as replacing legacy approaches like SNMP, it provides tangible benefits in terms of performance, scalability and visibility of the network. The session will also touch on how model driven approach is becoming the foundation for both network automation and visibility and how it transforms the operational environment of service provider networks.


Bio:


Rada Stanic is the Principal Systems Engineer, Software Solutions, Global Service Provider.  In this role, Rada is responsible for assisting Cisco’s Service Provider customers in developing their architectures and enabling new services that align with their business goals. Most recently, Rada’s key focus areas have been Software Defined Networking (SDN), Network Function Virtualisation (NFV), Programmability and Analytics. Prior to this role, Rada worked as an architect in the IoT space as well as all aspects of architecture, design and implementation of large carrier grade networks, spanning  areas of technologies such as MPLS, Metro Ethernet, Broadband, Mobility, Data Centre and Cloud. Prior to joining Cisco, Rada was a Senior Software Engineer at Alcatel-Lucent. Rada holds a Bachelor Degree in Telecommunications and Electronics Engineering.

WAND

Brad Cowie

Updates on the FAUCET project

APNIC

Geoff Huston

DNS, DNSSEC and DDOS


DNSSEC can offer authenticity to the DNS and prevent a number of attacks that attempt to inject false information into the name resolution system. However it is also possible to leverage DNSSEC to provide protection against DNS DDOS attacks. The essential change here is to allow DNSSEC-validating DNS resolvers to perform NSEC caching rather than simple NXDOMAIN caching. This talk will explain NSEC caching and why it can be a useful measure in our efforts to mitigate DNS DDOS attacks.

APNIC

Elly Tawhai

APNIC Updates

NZRS

Daniel Griggs

Accidental Data Analytics.


Bio: Daniel Griggs, After working in ISP operations for getting way too close to 20 years he thinks he's pretty close to getting it. Previously Daniel worked for a range of companies in a range of roles from unix systems administration, through to building and running a nationwide DWDM network. Recently Daniel has been working for NZRS updating their Anycast infrastructure, name servers and generally trying to be useful.


Abstract: While attempting improve the NOC traffic charts for DNS traffic, Daniel accidentally created a pretty good analytics platform that wasn't Hadoop based and doesn't require a large or even a small cluster. Another great outcome was it became easy to ask (and sometimes answer) questions about the traffic that was being received. In this talk Daniel will tell you how easy it was and encourage you to "give it a go". The main tool he'll be talking about is ClickHouse Database by Yandex, with a bit of Grafana to gussy it up.

ISOC

Aftab Siddiqui

Two years of good MANRS - Improving Global Routing Security
and Resilience

REANNZ

Jamie Curtis

Building Across the Pacific: A New International Network for
REANNZ and New Zealand New Zealand

When the new Hawaiki Submarine Cable Network comes online in mid 2018 it will be only the second cable system linking New Zealand across the Pacific to the rest of the world. The construction of the new system is a $400M investment bringing a new era of competition and service to Australia, New Zealand and Pacific Island nations. REANNZ has been responsible for managing the $15 million Crown investment into the project, funding provided over 10 years ago when the New Zealand government first signalled their intent to support the development of a new connection for the country.
In this talk we will discuss the impact this new cable will have on the research and education community in New Zealand, why it has taken 10 years to get to this point and what this means for all New Zealanders. We discuss our plans for how we will use our capacity on Hawaiki and how that can contribute to the global R&E network. Through our anchor tenancy, REANNZ has also had unique access to the design and construction process of the 14,000km (8,700mile) submarine system, and will share statistics, details and photographs of what goes into building a new submarine cable network. This talk aims to give insights into the fantastic engineering accomplishments involved in the submarine telecommunication cables that support the global Internet.

WAND

Shane Alcock

The Story of the Internet -- as told by the Application Layer

GCSB

Steve Martin

TICSA Update

NZ Police

Neil Tapp

Lawful Intercept Update

OpenLI

Dave Mill

Open Source Lawful Intercept WAND project

CFH

John Greenhough

An update on UFB and RBI projects

I will be touching on the current project status, and providing some insights into the RBI and MBS project design.

John Greenhough is CTO of Crown Fibre Holdings Limited, the Crown entity set up by the New Zealand Government to let and manage the contracts for building broadband networks in New Zealand population. As CTO, John is responsible for defining the technical standards for the networks, including SLAs, and ensuring these are met.


MTS

Andrey Dugin

Protect yourselves and others from DDoS


Every ISP technical guy knows that it is necessary to keep network infrastructure "clean" so you may not be a source of DDoS to you and others. But Internet investigation shows a lot of affected devices in the world. The most vulnerable infrastructure for use in DDoS-attacks are Asia Pacific technological leaders. Percentage of vulnerable infrastructure indicates country information security culture. Some technical analysis is provided for DNS/NTP/SNMP-amplification attacks. There are some standard and non-standard ways described in the presentation how to be unusable for DDoS-attacks type amplification.

Comments