Skip to content
Verified

On-Prem

What is programmable around a sales-led on-prem engagement, and how it is delivered.

Xuda On-Prem installs the same control-plane stack Xuda runs on its own regions onto infrastructure you control. It is a sales-led engagement, not a self-serve purchase, so the methods that carry the story, sizing a datacenter, provisioning it, keeping it patched, run from the dashboard and from Xuda's own operator tooling. What is reachable with an account API key today is the platform-topology read that sits underneath it.

Reading the platform's own topology

get_xuda_servers_info is the one public read of the fleet: id, countries, status and datacenter for every server Xuda currently runs in the live state (a server mid-deploy or failed is left out, so a shrinking result set during a rollout is not a decommissioned region). get_platform_servers_info answers the same shape of question from the platform side.

curl -s https://xuda.ai/cpi/get_xuda_servers_info \
  -H 'content-type: application/json' \
  -d '{"api_key":"YOUR_API_KEY"}'
const res = await fetch('https://xuda.ai/cpi/get_xuda_servers_info', {
  method: 'POST',
  headers: { 'content-type': 'application/json' },
  body: JSON.stringify({ api_key: 'YOUR_API_KEY' }),
});
const { code, data } = await res.json();
import requests

res = requests.post(
    'https://xuda.ai/cpi/get_xuda_servers_info',
    json={'api_key': 'YOUR_API_KEY'},
)
Cached for a minute

get_xuda_servers_info is cached for 60 seconds, two calls inside the same minute can return the same snapshot.

How an engagement is sized and delivered

get_account_datacenters (see Static Websites for the wider Account table) reads the datacenters already associated with an enterprise account, and get_region_hosting_options (see Cloud Hosting) reads the same region pricing a standard deploy uses. get_plans needs no credential at all, submit_support_ticket and the datacenter provisioning and replication methods (deploy_datacenter, configure_datacenter_replication, reinstall_datacenter_resources) run from a signed-in session, and the fleet-wide operator methods (ops_deploy_server, ops_batch_hotfix and the rest) check for an Xuda staff account before doing anything.

Methods