Skip to main content

Minutes

12-13 June 2017

The SEE 6/RIPE NCC Regional Meeting took place from 12-13 June 2017 at the Mediteran Hotel in Budva, Montenegro. A total of 111 attendees from 22 countries participated in the meeting, 33 of whom were from Montenegro.

Monday, 12 June 2017

Opening Plenary 09:00 – 10:30

Welcome

Serge Radovcic, RIPE NCC

Pavle Mijuskovic, The Agency for Electronic Communications and Postal Services, Local Host

Jan Žorž, Programme Committee Chair

Mad Men: About Selling Technology
Marco Hogewoning, RIPE NCC

Montenegro Internet eXchange Point (MIXP): A Success Story

Božo Krstajic, University of Montenegro

IPv6/BGP Plenary 11:00 – 12:30

NAT64/DNS64 Real Life Experiments, Warning and Also One Useful Tool
Jan Zorz, ISOC

Isolario - the Real-time Internet Routing Observatory

Luca Sani, IIT CNR

Daniel Karreberg, RIPE NCC, asked if Luca was familiar with the data provided by RIPEstat around BGP. He asked what differentiated Isolario from RIPEstat.

Luca replied that RIPEstat was not currently real-time and to use Isolario, you needed to open a BGP session and contribute. Isolario was not in competition with RIPEstat. As long as the public data increased, he thought this was a good thing.

MPLS in an IPv6-Only Network - Doable or a Bridge Too Far?
Berislav Todorovic, Juniper Networks

IOT Plenary 14:00 – 15:30

Concerns Related to the Security of the New Internet Connected Devices (ICD)
Goran Slavic, Serbian Open eXchange (SOX)

Panel: IOT Security and the Implications of a Lack of it

Moderator: Jan Zorz, ISOC
Panelists: Marco Hogewoning, RIPE NCC, Kevin Meynell, ISOC, Rok Arzenšek, NIL Communications, Goran Slavić, SOX

Daniel Karrenberg, RIPE NCC, said the IoT discussion mixed together a lot of things and it would be helpful if they could structure it into something that the community could deal with. Cutting users off was concerning and they didn't want to go down that route. The community's job was to secure and run the infrastructure. They needed to think about how they could do their job correctly so they didn't need to be regulated. Referencing an analogy to the airline industry, he said it was not only regulation that made an industry safe, but also culture. In the airline industry there was a security culture that incorporated a “no blame” analysis of incidents – where they had an anonymous voluntary reporting system. He said they should learn from this kind of culture and see if they could improve their image with those who were about to regulate them.

Paul Rendek, RIPE NCC, said IoT meant different things to different people. He still wasn't clear which venue network operators would use to talk about their piece of the puzzle. Maybe in the end this wouldn't be RIPE – they had talked with the European Commission, there was a group at the ITU as well. He asked where community members should go to address their portion of this debate.

Kevin Meynell, ISOC, said there was no single place. ISOC had performed an audit of IoT organisations and there were around 40 – some of which competed with one another, some of which that focused on different aspects.

Marco Hogewoning, RIPE NCC, said the community needed to carve out their niche. Many players in the IoT space tried to own it all, but this was impossible. They could see this at the national level, where telcos now had an opinion on cars. The RIPE community knew about networks – the End User device was less their problem. They should stick to what they could influence and what they needed to protect – their networks.

Paul encouraged those in the room to make some noise in order to own their slice of the debate. He encouraged them to support the RIPE IoT Task Force/Working Group and participate in it. The nice thing was that RIPE was respected as a community – they were approached by governments, the IEEE, GSMA and other bodies. The RIPE NCC was seeing people putting labels on the areas that they wanted to deal with – they had yet to see the operator community put a stake in the ground for what it wanted to handle.

Daniel said he wanted to address an earlier point that the RIPE NCC could evaluate End User devices. He said this was too big a task and not within their remit. It was the manufacturers who should do this. RIPE could inform them about minimum standards, but ultimately they should stick to the infrastructure and the handling of threats from these devices. IoT wasn't the only buzzword – another was “critical infrastructure” – and they were closer to that.

Florence Lavroff, Google, asked if RIPE should create an IoT WG.

Paul said there had been an IoT BoF at RIPE 74 and there had been a lot of interest – the room was so full that people were sitting on the floor. He said the community seemed to be interested in creating a WG.

IXP Plenary 16:00 – 18:00

IXP Database
Bijal Sanghani, Euro-IX

SOX Now, and Then

Nenad Krajnovic, Serbian Open eXchange (SOX)

Efforts to Establish a Neutral IXP in Albania

Daniele Arena, NaMeX, and Arjan Xhelaj, Academic Network of Albania (RASH)

IXP Panel - Challenges in the SEE region

Bozo Krstajic, MIXP, Nenad Kranjnovic, SOX, Angela Stanescu, InterLAN, Daniele Arena, NaMeX and Martin Semrad, NIX.CZ

Lightning Talks 18:00 – 18:30

Deploy360: The Future
Kevin Meynell, Internet Society

Sustainable Access: The Role of the Technical Community

Michael Oghia

Government as DDoS Facilitator
Dmitry Kohmanyuk, Hostmaster

BoF 18:30 – 19:30

Why Don't YOU Have IPv6?
Goran Slavic, Serbian Open eXchange (SOX)

Tuesday, 13 June 2017

Measurements/Infrastructure Plenary 09:00 – 10:30

Is Local Traffic Staying Local - What We See for the SEE Region with RIPE Atlas
Emile Aben, RIPE NCC

An audience member from Serbia asked if there was a plan for replacement of old probes that seemed to be a little unreliable lately. As a RIPE Ambassador in the Balkans, they were basically smuggling probes into the region around conferences – he asked if virtual appliances were being worked on as an alternative.

Emile said he didn't have a good solution for transporting probes and acknowledged that it was an issue – the RIPE NCC tried to come to local meetings and bring probes wherever possible. He said he would report this back.

Michael Oghia said he loved the IXP Country Jedi concept – the only problem was that he had no idea how to implement it. He wanted some kind of visualisation on the website that he could play with as a non-techie – he thought this would be effective for governments and related entities. Regarding RIPE Atlas credits, he had plenty that he didn't know what to do with, so he transferred them to other people. He said there must be a lot of unused credits out there and transferring these would be a good way to help people who needed them.

Emile said they were always happy to help people who needed credits. If people needed measurements started that were of general benefit, the RIPE NCC typically started them. He said they were currently working on making the Country Jedi nicer and easier for people to use.

Michael said that another idea might be to upload various screenshots for major IXP routes in places where people would be interested – even if it was something they could use in a presentation.

Daniel Karrenberg, RIPE NCC, thanked Michael for the good suggestions. He suggested that Michael find some allies and use the RIPE process to forcefully request these improvements. It would take some resources for the RIPE NCC to achieve this. He said it wasn't Emile's job to create a good user interface – but if there was a clearly voiced requirement from a lot of people, then the RIPE NCC could do this strategically rather than tactically. Regarding the earlier comment about software probes, he said the probes were relatively homogenous and predictable – so if you did measurements you could at least quantify the environment in which they are run. Once you virtualised these, you lost this aspect. They would have to categorise these measurements as being different. As soon as you did things like this, there was a reputation risk for the RIPE NCC— as you would be running this on much more powerful hardware and the RIPE NCC would be blamed if it was hacked. His final point was that the RIPE NCC had the resources to import RIPE Atlas probes legally. He wanted to see more of an effort to pay the import duty on behalf of the ambassadors, rather than going for a second class solution.

PC Based Software Router
Nenad Krajnovic and Goran Slavic, Serbian Open eXchange

RIPE NCC Web Services: Features and Updates

Adam Castle, RIPE NCC

DNS/NFV/IoT Plenary 11:00 – 12:30

RACI Talk: Software Defined Networking: Experimental Research on QoS
Slavica Tomovic, University of Montenegro

NETCONF and NFV

Rok Arzensek, NIL LTD

RACI Talk: IoT Turning Evil

Ivana Tomic, Imperial College London

Plenary 14:00 – 15:00

RIPE NCC Update
Axel Pawlik, RIPE NCC

Results from RIPE NCC Hackathons: IXP Tools and DNS Measurements

Vesna Manojlovic, RIPE NCC

Internet Market in Montenegro

Vitomir Dragas, Agency for Electronic Communications and Postal Services (EKIP)

Closing Plenary 15:30 – 16:15

.ME By Growing in Exports We Grow Our HO.ME
Natasa Djukanovic, .me registry

Farewell
Serge Radovcic, RIPE NCC