What a dedicated server changes for a streaming stack
A hypervisor deciding when your encoder gets a core is usually the first bottleneck a streaming setup hits, before the network is. A dedicated server has no scheduler above your own OS, so an encoder, origin or packager process gets whole cores instead of a slice of one, and a neighbor’s job on the same host cannot compete for them because there is no neighbor. See current configurations on our dedicated servers page.
Full-core encoding
A software encoder built on FFmpeg or GStreamer runs on cores nobody else touches, so a spike in another customer’s job elsewhere on a shared host cannot steal cycles from your encode.
Your own origin and packaging
Run an origin server such as nginx-rtmp or SRS, and a packager for HLS or DASH, on NVMe local storage rather than a shared volume, with the disk and network throughput dedicated to that one machine.
Test the path before you commit
Run ping and traceroute from five of our seven metros to your CDN edge or origin through the public looking glass, so you know what the path looks like before you provision anything.
What the platform runs on
- Windows Server and Linux one-click images
- Full root access, provisioned from the cloud console
- NVMe local storage for origin and packaging
- Hourly billing while you test a pipeline, monthly once it is live
Network and reliability
- AS55285 is our own network, across seven live metros
- 100% network uptime SLA: 5% of the monthly fee credited per hour of downtime we caused, up to the full month
- DDoS mitigation inline on the port, active by default
- Public looking glass to test the path yourself, from five of the seven metros
Private network from origin to edge
- Origin and packaging servers reach your edge servers over a private layer 2 network, off their public interfaces
- Viewers connect to the edge servers on their public addresses
- The private network can extend to your servers in our other datacenters, at no cost
Frequently asked questions
What ships from the catalogue for a streaming stack, how the network and SLA apply, and how to check a path yourself before ordering.
No. What ships from our dedicated servers page is single-tenant CPU bare metal that you run your own origin, packager or encoder on. Pair it with a CDN of your choice for edge delivery, or use it as the origin your CDN pulls from.
Test the path before you order
Looking glass • ping, traceroute • from five metros
Run a probe from one of our metros to your existing CDN or origin, then compare it with wherever your encoder sits today. Take a server by the hour if the difference is worth a live test.
Canadian and 
Dutch
