Cross connects at Nikhef: why speed and documentation matter

Nikhef, at Amsterdam Science Park, is one of Europe’s densest interconnection points. If your network peers in the Netherlands, there’s a good chance the path runs through here at some point. That density is exactly why a cross connect at Nikhef is so useful, and exactly why it’s so easy to lose track of one once it’s in.

What a cross connect actually is

Strip away the jargon and a cross connect is a physical run (fiber or copper) between two parties in the same facility. Your rack to a carrier. Your rack to a peer. A patch from your cage to theirs, terminated on both ends, carrying traffic. That’s it. There’s no magic in it; it’s a cable that someone has to physically route, terminate and label.

Because it’s physical, it’s also permanent in a way that logical config isn’t. You can’t roll back a cross connect from a console. Someone has to walk to the rack, find the right strand, and move it.

Where it goes wrong

The failure mode is rarely the install. It’s everything after. A cross connect gets ordered, patched and turned up, and then it just sits there. The person who ordered it moves on. The ticket gets closed. A year later there’s a patch panel with a dozen connects on it and no reliable record of which is which.

Now multiply that across years of growth. Unlabeled connects, undocumented endpoints, connects that might be live and might not. At that point nobody dares unplug anything, because the cost of guessing wrong is taking down a peering session or a customer link. So the panel just accumulates, and every future change gets slower and riskier.

Doing it right

The fix isn’t complicated, it’s just disciplined. Every connect gets labeled (per your scheme, not ours) so it matches the rest of your documentation. Photo proof per connect, so there’s a visual record of what’s actually patched where. And the connect is recorded in Netbox, so the documentation lives somewhere durable rather than in someone’s memory.

Do that consistently and the hard problems get easy. A migration becomes a matter of reading the records instead of tracing cables by hand. A clean-up becomes safe, because you can tell at a glance what’s live and what’s abandoned. The work is the same either way. The difference is whether it’s traceable afterwards.

Why have hands at Nikhef

When networks peer, being patched fast matters. A new connect that’s quick to turn up means a new peer or carrier comes online when you need it, not weeks later. So speed counts.

But the same partner documenting everything counts more over time. A fast install you can’t find again is a future problem you’ve paid to create. Having on-site hands at Nikhef who both patch quickly and record what they did (labeling, photos, Netbox) means today’s connect doesn’t become next year’s mystery cable. We’re already inside the facility doing exactly this, including installs, removals, migrations and console sessions for kit you already have racked here.

Got connects at Nikhef you’d like documented properly, or new ones to turn up? Get in touch and we’ll take it from there.