A blended upstream mix and direct peering, handed off from 1G to 100G. Your policy stays yours; ours is documented, exported and yours to read.
One hand-off, multiple upstreams and peers. Routing decisions follow documented preferences rather than whichever session came up first.
Bring your ASN, communities and import/export policy. We publish our filters and community list so you can predict exactly what arrives.
We test upstream withdrawal rather than trusting it. The behaviour, including the unattractive parts, is documented for your service.
| Hand-off | Suits | Interface |
|---|---|---|
| 1 Gbps | Single services, game and voice platforms | RJ45 / SFP |
| 10 Gbps | Hosting platforms, CDN edges | SFP+ |
| 25 / 40 Gbps | Growth and aggregation paths | SFP28 / QSFP+ |
| 100 Gbps | Transit aggregation at scale | QSFP28 |
Commit, burst and multi-port bundles are priced per requirement — tell us the 95th percentile you expect, not the peak you once saw.
Documented local preference per prefix class, adjusted with communities where a customer needs something other than the default path. You can see the policy rather than infer it from traceroutes.
Yes — originate on AS209101 if you prefer, or bring your own ASN and we will build the session and filters with you.
Traffic withdraws to the next preferred path. We exercise that transition deliberately, and the expected convergence behaviour is part of your service documentation.
Settlement-free where traffic is reasonably mutual and your objects are in order. If it is lopsided, we will say so and talk about it openly rather than sulk in a filter.
95th percentile by default, from interface counters you can also poll yourself. Burst and commit terms are written into the quote.
Ports and commits, address space, racks, or a routing question you want a second opinion on. You get an answer from an engineer, with numbers attached.
