Instance families
XenonCloud offers a small, well-curated set of instance types designed around common workload patterns:
- Launch – low-resource workloads, CI workers, hobby projects.
- Scale – most production APIs, web applications, and microservices.
- XL & specialized – CPU-intensive work, background jobs, and analytics.
Sizing guidance
Start from your workload profile:
- CPU-bound: choose more vCPUs before adding RAM.
- Memory-bound: prioritize RAM; add CPU when you see ready queue growth.
- IO-bound: consider high-IO storage or moving heavy queries to dedicated nodes.
Instance limits
New accounts start with conservative limits to protect both you and the platform:
- Maximum instance count per project.
- Maximum vCPUs and RAM per region.
- Snapshot and volume capacity caps.
These can be raised quickly through a support request that includes your expected usage pattern.
Recommended patterns
- Avoid running unrelated workloads on the same instance; use multiple small instances instead.
- Use at least two instances for production services behind a load balancer or reverse proxy.
- Separate stateful services (databases, queues) from stateless app servers.