IXCDN Node Onboarding

Confirm resources, network and facility readiness before deployment

IXCDN is not an open PCDN scale-out network. Binaries, node claiming, acceptance and routing are unlocked only after servers, bandwidth, GPUs, facilities and operational evidence are approved.

Flow

Onboarding flow

01

Prepare evidence

Confirm subject, facility, network, GPU/server, bandwidth and auditable materials.

02

Apply

Submit the resource application for CDN, AI token, GPU inference or hybrid workloads.

03

Review

Compliance, resource authenticity, regional value and operational continuity are reviewed.

04

Deploy

Approved partners unlock em-edge / em-edge-ai binaries, templates and node claiming.

05

Acceptance

Version, health, connectivity, load, logs and declared resources are checked.

06

Routing

Accepted nodes enter routing pools with status, load, traffic and earnings reconciliation.

Resource

Server and AI node requirements

Resources can be onboarded in phases, but declarations must be accurate and verifiable. GPU nodes are prioritized by inference capacity, stability and secure configuration.
  • Linux x86_64 host with root/sudo operations, time sync and basic monitoring.
  • Edge cache nodes should have SSD/NVMe cache storage, stable public IPs and sustained bandwidth.
  • AI nodes should declare GPU model, memory, driver/container runtime, inference method and workload type.
  • Upstream model or API credentials are not submitted in public forms; configure them on the node through the deployment security flow.

Network

Network node requirements

IXCDN values controllable, low-detour and reconcilable network resources, avoiding compliance and settlement risks from weakly managed PCDN nodes.
  • Provide region, carrier, ASN/BGP, IX/Peering, IP ranges, IPv6 support and uplink bandwidth.
  • Declare cross-region or cross-border limits, egress cost, packet loss, latency, peak bandwidth and availability windows.
  • Support precise routing by region, carrier, load, health and service policy to reduce cross-region traffic and settlement cost.
  • Do not use residential proxies, hidden PCDN, unclear IP sources or unauditable forwarding resources.

Facility

Facility and operations requirements

Facility quality affects uptime and acceptance speed. Complete evidence helps determine whether the node can join high-value workload pools.
  • Submit facility address, tier, rack count, install lead time, photos or verifiable evidence links.
  • Describe power redundancy, UPS/generator, cooling, rack power, power connectors and PDU capability.
  • Provide 24x7 NOC, remote hands, incident SLA, inspection and change-window capability.
  • After launch, support health checks, upgrades, incident response and earnings reconciliation.

Deploy

Deployment steps

After approval, the affiliate portal shows downloadable binaries, node tokens, configuration templates and installation instructions. The commands below are illustrative and do not replace the portal instructions.
Download binary

Download the em-edge / em-edge-ai package for the right architecture and workload type.

Configure node

Set node token, cache paths, listen ports, log paths and GPU/inference service connection.

Start service

Install the systemd service or container unit, then confirm startup and log collection.

Health checks

Verify ports, connectivity, disks, GPU, load, version and reporting links.

Wait acceptance

After launch, wait for acceptance. Traffic ramp-up follows business routing policy.

Install command example
# Actual package, token and config path are shown in the affiliate portal
sudo ./em-edge install --config /etc/ixcdn/em-edge.toml
sudo systemctl enable --now em-edge
sudo systemctl status em-edge

Guardrails

Do not before launch

  • Do not promise IXCDN traffic or earnings before approval.
  • Do not submit false bandwidth, IP, GPU, facility photos or unauditable logs.
  • Do not put upstream model keys, customer credentials or sensitive secrets in public forms.
  • Do not use PCDN, proxy pools or residential networks without clear source and control.

Portal

What becomes visible

Node status

Online state, version, health, alerts and acceptance status.

Resource load

CPU, memory, disk, GPU, bandwidth and inference service load.

Traffic reconciliation

Requests, hits, origin fetches, DNS/GSLB, anomaly checks and logs.

Earnings view

Reviewed earnings for CDN, edge security, AI token and GPU inference workloads.

Ready to join? Submit resources first

Prepare subject, facility, network, GPU/server and NOC materials, then submit the node resource application. Approval unlocks download, deployment, acceptance and routed earnings.