SURF.net news

Tag: SURFnet8

  • Ready for the future with an 800G peering network

    Ready for the future with an 800G peering network

    SURF is building a new generation of the SURF network. This blog is part of a series of blogs about SURFnet Infinity. With SURFnet Infinity, SURF has opted for a vendor-neutral architecture based on open standards, ensuring that the network can continue to develop without restriction. You can read more about the new SURF network here.

    In this blog, we’ll take a closer look at the technology and focus on SURFnet Infinity’s new peering network.

    The first step

    In 2025, SURF organised a mini-competition in collaboration with the broker Nomios. Juniper-HPE emerged as the winner for the core and peering domain with its PTX10000 platform. In addition to the hardware, the anti-DDoS solution used by SURF was also part of this mini-competition. As part of the Juniper-HPE proposal, the new software package from supplier Corero, Corero Smartwall One, was selected. This version has been specifically developed to support Juniper PTX. This new version of the software package offers a clear upgrade path from the existing Corero implementation currently used by SURF.

    Replacement of the peering routers

    In 2025, we quietly carried out a migration of the existing peering routers in the SURFnet8 domain. Until August 2025, all of SURF’s external connections were directly connected to the core routers in Amsterdam. There was a need for greater capacity and increased resilience within the SURF network. As the high-level architecture of the SURFnet Infinity network had already been finalised at that point, we decided to apply the new architecture here for the first time. Dedicated peering routers enable faster capacity expansions, a higher level of resilience in the event of DDoS attacks, and a smaller blast radius. The latter is a key aspect in the separation of network services, as it allows us to significantly limit the impact on other domains.

    In early 2026, we migrated the existing Corero anti-DDoS service to the new package, Smartwall One. This is still linked to the existing Juniper MX-based peering routers, but offers plug-and-play capability for integration with PTX routers.

    In July and August 2026, the migration to the new PTX platform took place. New PTX10002-36QDD routers were installed at the main peering locations, Equinix AMS7 and Nikhef. Both routers were provided with direct connections to the new SURFnet Infinity core network, which is currently under construction, as well as direct connections to the SURFnet8 core network. This enables us to gradually migrate customers from SURFnet8 to SURFnet Infinity, with sufficient capacity and without any impact on internet services. By migrating the peering routers ahead of the metro sites and connecting them to both core networks, we can focus fully on migrating the customer sites. There is no need to take routing or capacity into account, as this complex work has already been carried out.

    A diagram is shown below:

    Migration process

    From a bird’s-eye view, such a migration process looks like this:

    • – installation of the new routers
    • – establishing connections with the SN8 core and the SN∞ core
    • – integrate anti-DDoS platform
    • – Migrating the external connections one by one

    This migration was preceded by a major project plan, with the constant aim of causing no impact – or as little as possible – on service delivery. A great deal of planning went into this, particularly at our Nikhef site. Within SURFnet8, one of SURF’s most important peering locations was Digital Realty AMS9. In 2025, SURF decided to relocate the network core to the data centre of one of its members, Nikhef. The Nikhef data centre is one of Amsterdam’s most important connectivity hubs, making it the ideal location for peering. More information about the collaboration: https://www.surf.nl/en/news/surf-moves-network-core-from-private-business-to-member-organisation

    Following consultations with various external parties and peering partners, the necessary cross-connects were requested between Digital Realty AMS9 and Nikhef. This enabled us to connect existing links from parties not physically present at Nikhef to the new peering nodes.

    Field engineers will then migrate all connections from the MX peering router to the PTX peering routers. This means that the optics and fibres will be reconnected on a one-to-one basis. Afterwards, network engineers can use the automation platform developed by SURF to transfer the configuration at the touch of a button. This helps to ensure that the migration proceeds smoothly and without errors.

    Actual migration: a shift in traffic patterns

    The actual migration began in the week of 20 July 2026. We opted for a two-stage approach. At the Asd002a site, the migration was carried out using the ‘big bang’ method. On Monday 20 July at around 10.00, the first gateways were migrated to the new PTX platform. The entire migration was completed that same day at around 13.30. The graph below clearly shows the shift in traffic from SURFnet8 peering (in red) to SURFnet Infinity∞ (in green) at the asd002a site.

    Shift in traffic from SN8 MX peering routers (in red) to SN∞ PTX peering routers (in green) at site ASD002A

    For the peering migration at site Asd001b, we opted for a phased approach. Various external connections needed to be relocated from site Asd001b (Digital Realty AMS9) to Asd066d (Nikhef). Due to the need to re-establish several campus cross-connects and set up new connections, this migration has been spread over several weeks. We are currently still awaiting the final handover and hope to have fully migrated to the PTX platform at Asd066d by the end of October 2026.

    Shift in traffic from SN8 MX peering routers (in red) to SN∞ PTX peering routers (in green) at location ASD001B/ASD066D

    Distribution of traffic between SURFnet 8 and SURFnet∞

    At present, the migration to the SURFnet Infinity metro network is still well into the design and installation phase. For most customers, this means that internet traffic is still routed via the SURFnet 8 core network. For a small number of metro locations, the migrations to SURFnet∞ have already been completed. We can also already see that traffic from the peering nodes is travelling via the SURFnet∞ core network. This is shown below, where traffic from our peering routers to SURFnet Infinity is shown in green and traffic to the SURFnet 8 core network is shown in red.

    Initial traffic to customers on the SN∞ network (shown in green) by peering location.

    Next step: 3rd peering PoP in Frankfurt

    To make the SURF network even more robust, we have decided to also implement geographical redundancy. As part of the expansion of our Cross-Border network, we are also building a new PoP in Frankfurt. This will enable us to establish external connections here as well, as a fully-fledged part of our peering strategy. The aim is to establish connections with other networks at this location later in 2026, thereby increasing our redundancy and capacity.

    Further information

    Would you like to find out more about SURFnet Infinity?

    Follow this project on this page. We also post regular project updates on the network dashboard. Alternatively, please contact us at joachim.opdenakker@surf.nl

  • Highlights from Internet2 TechEx 2025

    Highlights from Internet2 TechEx 2025

    At Internet2 TechEx 2025, global developments in research and education (R&E) network infrastructure were high on the agenda. In a well-attended update, Brenna Meade (International Networks, Indiana University) shared an overview of key steps being taken worldwide to scale capacity, resilience, and automation in federated network services.

    Major capacity upgrades, including transoceanic links

    Meade outlined a broad range of ongoing and planned upgrades across international backbone and exchange infrastructures. This included new 400 Gbps transoceanic links—critical for data-intensive research and collaboration across continents. She also highlighted activity and continued evolution across multiple Global Exchange Points (GXPs), including FUJI-XP, SOE, GOREX, MANLAN, MOXY, NetherLight, and Pacific Wave. Together, these hubs form an important foundation for high-performance global connectivity between R&E networks.

    NSI reaches production readiness

    A key milestone highlighted in the session: NSI (Network Service Interface) has reached production readiness. NSI enables interoperable, automated service provisioning across network domains. In practice, this supports standardized ways for organizations and networks to request, set up, and manage end-to-end services across multiple administrative boundaries.

    For NRENs, this aligns closely with the push toward scalable, federated connectivity: less manual coordination, faster delivery of services, and more reusable interfaces and operational agreements between domains. Reaching production readiness is therefore a concrete step toward more automated and dependable international network service ecosystems.

    Technology—and the community behind it

    Beyond the technical program, TechEx continues to stand out as a strong community meeting point. Informal conversations between sessions, sharing experiences across very different operating contexts, and social traditions such as the 5K fun run all reinforce the trust and relationships that are essential to building and operating resilient infrastructure.

    In summary: TechEx 2025 underscored how global R&E networks are moving forward on both capacity and automation—with NSI marking a notable step toward interoperable, federated service delivery.

    Screenshot

  • SURF partners with HPE and Salumanus for multi-vendor and high-performance network design

    SURF has selected HPE to deliver solutions for the backbone and border network domains, and Salumanus is selected to deliver the coherent transceivers. Together with partner Nomios we will implement the next generation SURF network infrastructure, designed to meet the rapidly growing demand for bandwidth, predictability, and resilience across the Netherlands’ research and education community.

    To be able to meet future challenges, SURF needs to innovate their network continuously. For that purpose, it has designed a new network architecture and will start replacing the network in various phases over the next few years. Within the new architecture, the SURF network delivers 800GE, 400GE and 100GE connectivity. Using OpenZR+ technology from Salumanus, SURF increases its backhauling capability, both in terms of bandwidth and distance.

    “This new architecture gives SURF exactly what our members need: higher capacity, predictable performance, stronger security and the freedom to scale without being tied to a single vendor. It allows us to upgrade faster, operate the network more efficiently and ensure that every institution benefits from a platform that is ready for the demands of the next decade.” Said Harold Teunissen, Head of Network & Infrastructure Services, SURF

    HPE Networking for backbone & border

    HPE delivers the backbone and border domains of the new SURF network with PTX Series Routers (PTX10008 and PTX10002-36QDD, with Express 5 silicon’s unparalleled power efficiency), connecting SURF’s IP peering and the NetherLight global exchange point. The solution includes HPE vRR (Networks virtual Route Reflector) for simplified routing management and scalability, as well as Corero comprehensive anti-DDoS capability to safeguard research and education traffic against large-scale cyberattacks—ensuring resilient, efficient and secure connectivity for critical scientific and academic applications.

    “HPE is proud to support SURF in building a resilient, high-performance network for the Netherlands’ research and education community. By delivering advanced routing, security, and data privacy solutions, we help empower local institutions to collaborate, innovate, and protect sensitive information”, commented Dobias van Ingen, EMEA CTO HPE Networking

    Salumanus technology  

    At the heart of this transformation is the migration of SURF’s existing CWDM-based access network to DWDM (Dense Wavelength Division Multiplexing) technology. This step is crucial to enable a smooth and scalable upgrade of existing links to 100G and higher capacities, ensuring that member institutions can benefit from next-generation optical performance.

    In parallel, the backbone network will be upgraded to 400G and 800G using IP over DWDM technology. This approach allows SURF to deliver ultra-high-speed connectivity, eliminating the need for separate transponder hardware, minimizing power consumption and space utilisation while significantly reducing capital and operational costs.

    “We are proud to support SURF in its mission to deliver world-class connectivity, our OpenZR+ technology and DWDM solutions will help build a network that is smarter, more efficient, and ready to meet the demands of the future.” Said Marcin Bala, CTO of Salumanus

    Continued collaboration with Nomios

    Nomios plays a key role as systems integrator in coordinating hardware purchasing and lifecycle management across suppliers and domains for the next generation of SURF network infrastructure. SURF has engaged Nomios to assist in selecting technology partners and executing the mini competition process on behalf of SURF. This collaborative model allows SURF to continuously innovate while maintaining maximum flexibility and vendor diversity.

    “We’re supporting SURF in selecting the right technology mix and keeping the multi-vendor model practical across all network domains. This next step in their architecture gives the research and education community in the Netherlands the bandwidth and predictability they need for the years ahead.” Said Mohamed El Haddouchi, Managing Director Nomios

    About SURF

    SURF is the IT cooperative of Dutch education and research institutions. The members, the owners of SURF, join forces to develop or procure the best possible digital services, work together on complex innovation issues and develop and share knowledge with each other.

    SURF designs, builds, and manages the national research and education network, the SURF network, that connects universities, research institutes, and other knowledge organisations.

    About Salumanus

    Salumanus provides advanced optical networking solutions across Europe, specialising in DWDM and transceiver technology. The company supports network operators, enterprises, and research networks in building flexible and cost-effective infrastructures.

    About Nomios

    Nomios is a European leader in secure networking and cybersecurity services and solutions. The company designs, implements, and manages resilient digital infrastructures, with a strong focus on innovation and long-term collaboration.

  • SURF and Nomios collaboration for a more flexible and innovative network

    For the renewal of network equipment, SURF has awarded the tender to network supplier Nomios. Through Nomios, we can purchase network equipment from different suppliers, which strengthens the innovative power of our networks. This allows us to respond more flexibly and efficiently to the network needs of educational and research institutions in the Netherlands.

    With Nomios as the intermediary, we do not replace the SURF network all at once, as we did with previous generations, but gradually renew the network where the need is most urgent. This approach gives us the flexibility to choose the most appropriate network equipment and supplier per domain or network layer. This allows us to optimally match the needs and requirements of educational and research institutions. By partnering with Nomios, we are moving towards the next generation of the network, SURFnet∞ (SURFnet Infinity), which will eventually replace the current SURFnet8. 

    Increased bandwidth and network solutions

    The new contract with Nomios applies to both the SURF fibre-optic network and the SURF data centre networks. On SURF’s national network, member institutions require more bandwidth every year. Also at the SURF data centre networks, which for instance support the national supercomputer Snellius, the demand for bandwidth is increasing. In addition, the SURF network and data centre networks are used for large international research projects, such as the Large Hadron Collider, Square Kilometer Array and the Einstein Telescope, which not only demand a lot of bandwidth but also place specific demands on the network.

    Benefits of flexible development

    The flexible development of our network offers tangible benefits to all member institutions: 

    • Optimal use of existing resources: by replacing equipment in phases, we extend the lifespan of our current infrastructure.
    • Better adaptation to change: we can respond faster to unexpected challenges and technological developments from suppliers. 
    • Direct access to innovation: we can acquire new and innovative technologies quicker and apply them to the SURF networks. 
    • User-friendly optimisation: with the diversity of network equipment available, we can optimise the network based on the broad requirements of education and research institutions. 

    Think about SURFnet Infinity

    In the coming months we will work with our partners and members to further develop SURFnet∞ (SURFnet Infinity), the new generation of the SURF network. During the knowledge session Procurement, architecture and design SURFnet Infinity on the 21st of November, we will present the first draft version of the network. We invite all stakeholders to think along and participate.

    About Nomios

    Nomios is a leading European provider of cybersecurity and secure networking solutions with broad experience in various sectors, including education and science. Nomios offers a wide range of services, including professional and managed services, support and SOC teams, making it a strong partner for SURF and its members.

  • SURF stops support for IP Multicast

    Since the nineties, SURF has supported IP Multicast within its network. What once started as a special network for this technology, grew into a standard part of the SURF network. However, after decades of use, SURF has decided to phase out IP Multicast. In this blog post I will explain what IP Multicast exactly is, how it was used within SURF in the past, and why the technology must now make way for modern alternatives.

    What is IP Multicast?
    IP Multicast is a technology that makes it possible to send data from a server to multiple recipients in a scalable way. The server only needs to send the data once, after which the network takes care of the replication to all recipients.

    An example of this is the broadcasting of a live lecture to various universities in the Netherlands. Instead of the live video stream being sent multiple times from the source (once for each recipient), it is only sent once via IP Multicast. The network then ensures that the stream is distributed to all affiliated universities. This reduces the required bandwidth and ensures more efficient use of network resources.

    Use and applications within SURF
    Over the years, this technology has been tested and used in various ways by affiliated institutions. Examples of applications were:

    • Sending television channels to student houses.
    • Broadcasting webcam images, such as those of the coast of Vlissingen.
    • Sending satellite weather images from Germany to the KNMI and universities in various European countries.

    Why is IP Multicast used less these days?
    Although IP Multicast once held great promise, it is increasingly being used less and less due to management complexity and limitations in modern networks. In cloud environments, where networks are often dynamic and virtual, IP Multicast can be difficult to implement. Many networks prefer unicast streaming and Content Delivery Networks because of their greater scalability and flexibility.

    Unicast streaming sends separate data streams to each recipient, which requires more bandwidth, but with the increased capacity of modern networks this is no longer a major issue

    Why is SURF discontinuing IP Multicast?
    None of the old IP Multicast applications are still active, and IP Multicast has not been used on the SURF network for some time. Supporting IP Multicast introduces significant protocol complexity, both for SURF and for the connected institutions.

    By phasing out IP Multicast, SURF is simplifying its network protocols. This decision marks the end of an era, but also opens the door to new technologies and enables a less complex migration to a next-generation network in the future.

  • New names for SURF network services

    As of January 2024, the names of the network services will change. We are doing this to better align our services with the terms and conventions used in the network community. We hope that this will make it clearer to network administrators which services are provided.

    Current service nameNew service name per 1 Jan 2024
    SURFlichtpadEVPN – Point-to-point
    SURFlichtpad – RedundantEVPN – Point-to-point Redundant
    L2VPNEVPN – Multipoint
    L3VPNL3VPN – Multipoint
    SURFinternetInternet
    Multi Service Poort (MSP)Service Port

    Lightpath becomes EVPN

    The term lightpath, although well established in SURF language, has always caused some confusion. This is not an optically illuminated path but a dedicated network connection over the service layer between two points. EVPN (Ethernet-VPN) is what is provided here and the addition (Point to point vs Multipoint) indicates whether it is set up between two points or more than two points.

    Multi Service Port becomes Service Port

    Under the new All-In Network Tariff, which came into effect on September 1, 2023, institutions can adjust their network port configuration as they see fit. This effectively makes all the network ports “multi” service ports. That is why they will simply be called Service Port from 2024.

    The changed names have been implemented in the network dashboard, on (mijn)surf.nl and on invoices.