BGP transit for your own AS.
From AS-SET to session, with a policy you can read up front.
Xyphen IT delivers IP transit for organisations that depend on stable, predictable connectivity. We have the expertise to build a reliable, flexible solution that grows with your network and capacity requirements. Direct routes, low latency and capacity that moves with your infrastructure.
Network performance sets the pace.
Applications, platforms and services demand constant availability and predictable network behaviour. IP transit is the foundation on which traffic is handled quickly, reliably and efficiently. With direct upstreams, low latency and scalable capacity, Xyphen IT delivers a transit solution that fits the technical requirements of your environment — from data centre and colocation to business-critical office infrastructure. Our IP transit services are set up based on your specific situation.
Our approach to IP transit goes beyond just delivering capacity. We provide the technical knowledge and operational support needed to keep network traffic structurally manageable, scalable and performant.
This is what a session with AS211588 looks like.
No marketing layer in between: you set up a BGP session to our network and receive the full table, including large communities and with BFD on the session. Below is an example configuration and what an established session looks like. Multihoming, prefix limits and MD5/TCP-AO are agreed during the intake.
! BGP session to AS211588 (example, Arista EOS) router bgp 64512 neighbor 45.83.205.1 remote-as 211588 neighbor 45.83.205.1 description XYPHEN-IT:TRANSIT neighbor 45.83.205.1 send-community standard extended large neighbor 45.83.205.1 bfd neighbor 45.83.205.1 maximum-routes 1500000 ! address-family ipv4 neighbor 45.83.205.1 activate ! IPv6 session: same setup, with the v6 neighbor from your intake
peer-rtr# show ip bgp summary Neighbor V AS MsgRcvd Up/Down State PfxRcd 45.83.205.1 4 211588 4821337 127d04h Estab 1024893 peer-rtr# show bfd peers | inc 45.83.205.1 45.83.205.1 Up 300ms x3 Ethernet1 peer-rtr# show ip bgp 45.83.205.0/24 | inc Large Large Community: 211588:1:10 211588:2:1103 peer-rtr# traceroute 9.9.9.9 1 core1.ams.xyphen-it.net (45.83.205.1) 0.6 ms 2 nl-ix.core 1.1 ms 3 dns9.quad9.net (9.9.9.9) 1.4 ms
Handoff from 1GE up to 100GE, IPv4 and IPv6 dual-stack on the same session or separated — however you've set up your network.
The technology your session runs on.
IP transit from AS211588 runs on infrastructure we build and manage ourselves: leased dark fiber that we light with our own DWDM equipment, routing on Arista, Juniper MX and Huawei, and an EVPN/VXLAN backbone between our Amsterdam core locations. We take transit from Fusix Networks, Eranium, NL-ix and Worldstream; in addition, we have seven connections to internet exchanges — a mix of 10G and 100G pipes.
- MP-BGP
- Dual-stack IPv4 + IPv6
IPv4 and IPv6 on separate sessions or combined via MP-BGP — however your network is set up. No IPv6 space of your own? As a transit customer you can request an IPv6 prefix from us, so you can still go dual-stack onto the internet.
- RFC 8092
- Large communities
Traffic engineering with 211588:*:* large communities: prepends towards specific upstreams or exchanges, selective announcements and informative tags about where a route was learned. You'll receive the full community list at delivery.
- RFC 5880
- BFD
Sub-second detection of link failures, so failover doesn't have to wait for BGP hold timers. Available as standard on every session.
- TCP-AO
- Session security
MD5 on the session, also visible in your dashboard, or TCP-AO on request. Max-prefix monitoring on both sides and GTSM where desired, agreed during the intake.
- IRR
- Filtered announcements
Our announcements are filtered on AS211588:AS-XYPHEN; from your AS we expect an up-to-date AS-SET and a correct PeeringDB entry.
- RIPE
- LIR services
Through our parent company Innovation Harbor B.V. we are a RIPE NCC member. Alongside transit we offer LIR services such as AS sponsorship and guidance with resource requests — useful if you want to set up your own AS or address space.
- DWDM
- Our own foundation
No resold waves or rented routers: the optical layer, the network and the NOC are ours. We resolve outages ourselves, without an intermediary.
What we accept from you, in this order.
Every customer session gets the same import policy, term by term. RPKI validation comes before it: a route with an invalid ROA goes no further. The order matters, so it is listed here in the order the router reads it.
- 01gshutGraceful shutdown
Routes with community 65535:0 get local preference 0 and continue through the policy. This lets you drain a session cleanly for maintenance (RFC 8326).
- 02rtbhBlackholing per address
A /32 or /128 within your prefix list with 65535:666 or 211588:0:666 is dropped at our network and passed on to exchanges and transit. This term deliberately comes before the length filter.
- 03bogonsBogons
Address space that does not belong on the internet, such as private, reserved and documentation space, is rejected.
- 04own-spaceOur own space
We do not accept AS211588 prefixes from a customer.
- 05too-longToo specific
Longer than /24 for IPv4 or /48 for IPv6 is rejected, apart from the blackhole route.
- 06as-pathAS path
The path must belong to your AS. If you have downstream customers with their own AS, we include them during the intake.
- 07prefixesYour prefix list
What is in your prefix list, and more specifics up to the length limit, is accepted, tagged as a customer route and gets local preference 200.
- 08denyEverything else
Anything not accepted above is rejected.
Changes to your session or policy are made with commit confirmed: after the change we check your session and the accepted prefixes, and if something is wrong, the router rolls the change back by itself within five minutes.
Why network configuration belongs in version control →How your session comes about.
- Intake and checksAS number, AS-SET, location, port and what you want to receive from us. We check your AS-SET in the IRR, your PeeringDB record, the RIPE holder of your space and your ROAs as our routers see them.
- Prefix listBuilt from your AS-SET and guarded against your PeeringDB limit. If the list later changes a lot, you approve the new version in the dashboard first.
- Session and policyThe session goes into our network configuration under version control, with the standard import policy below. What is on the router is also in the repository.
- HandoverWe check that the session comes up and accepts your routes. From then on you follow its status yourself in the dashboard.
You receive the session details by mail and in your dashboard; an MD5 key reaches you separately, never in the same mail.
IP transit suits organisations that:
- Run their own networks or autonomous systems
- Require low latency and stable performance
- Are looking for scalable internet connectivity
- Want strategic and technical control over network traffic
Technology that earns trust.
- Deep knowledge of routing, BGP and network infrastructure
- Direct contact with engineers
- Solutions tailored to real-world practice, performance and manageability
- Integration with colocation and network architecture
IP transit at Xyphen IT is about control, speed and continuity.
Get a network that moves with you.
Whether it's about growth, optimisation or stability: we make sure IP transit contributes to your performance, control and clarity.