Four cloud services, one contract
Managed Public Cloud covers four practical needs under a single commercial and support relationship: protected object-storage backups, Linux cloud instances, high-availability PostgreSQL or MySQL, and managed Kubernetes with its own private container registry. Every order is a specific, named package rather than an open-ended pay-as-you-go account, so you always know what you are committing to and what it costs. Two options are intentionally not offered for new orders because we don't think they hold up commercially yet for our customers: open-ended AI inference budgets, and single-node ('Essential') databases with no failover.
How pricing works
Every plan is shown under a clear Silurian name with one current, all-inclusive monthly price - Silurian checks and updates pricing regularly so what you see is accurate, and you never need to track a separate bill or currency conversion yourself.
Where your data lives
Available locations depend on the service: protected 3-AZ (three-availability-zone) backup is currently offered in France; cloud instances and high-availability databases are available in France, Germany and Poland; managed Kubernetes with its private registry is available in France and Germany. Whichever region you choose is recorded on your order and re-verified during provisioning - it cannot be silently changed afterwards.
A monthly price with no surprises
Every price shown is a net monthly amount before tax. Backup capacity has a clear stated limit. Hourly-billed compute and Kubernetes control-plane components are converted into one fixed monthly package assuming continuous use (744 hours), so you see a real monthly number instead of an hourly rate you have to calculate yourself. If you need more capacity or a different architecture later, that's a plan change you request and approve - your bill never grows on its own.
How ordering and provisioning work
Paying for a cloud service starts a provisioning review rather than creating an unchecked resource immediately. Silurian verifies current cost and capacity in the selected location, confirms the intended use and creates the contracted topology. Record the delivered resource name, region, addresses, quotas and recovery contacts before deploying a workload.
Operating responsibilities
For instances, you control power, operating-system access, applications and data. For managed databases, you control schemas, users and application connections while the platform operates the database nodes and failover layer. For Kubernetes, you control workloads, policies and application data while the control plane is managed. Backup-storage credentials provide access to a target; they do not configure or verify the backup job itself.
Backup and object storage
This service uses S3-compatible object storage replicated across three availability zones for resilience. Your package includes a daily backup workflow, a stated retention period and restore assistance from Silurian - but you still need to point your backup source at it, set your credentials, and confirm your first backup actually completed successfully. Capacity, retention and what's covered by restoration help are exactly what's stated in your order.
Cloud instances
Every instance comes with a public IPv4 and IPv6 address included. Development-tier ('Discovery') instances use shared resources and suit testing, sandboxes and development work; production-tier ('General Purpose') instances get guaranteed, non-shared resources. Floating IPs, snapshots, extra block storage volumes, load balancers, Windows licences and operating-system management are add-ons, not automatic inclusions - check your order or ask us if you're unsure what you have. If you just need one simple, economical server, a standard VPS is usually the better and cheaper fit than a cloud instance - ask us which suits your project.
High-availability databases
New managed MySQL and PostgreSQL databases start on a two-node 'Business' topology: automatic failover, a 99.9% uptime SLA, daily encrypted backups kept for 14 days, and point-in-time recovery. You can upgrade storage, instance size or plan tier later, but you cannot downgrade an existing database's plan. For a small or non-critical database, Managed Web Hosting or a self-managed VPS database is usually more cost-effective - talk to us before committing to high-availability if you're not sure you need it.
Managed Kubernetes
Production Kubernetes packages include a Standard control plane, General Purpose worker nodes and a private container registry. The Standard control plane runs dedicated etcd with a 99.9% SLA in a single availability zone, or 99.99% where 3-AZ regions are available. Load balancers, persistent volumes, deploying and monitoring your own applications, and backing up what runs inside the cluster are all separate from the base package - the control plane and workers being managed does not mean everything running on top of them is automatically covered.
Support and responsibility
Silurian is your single point of contact for billing, invoicing and support across every Managed Public Cloud service - hardware and platform-level incidents are handled by Silurian's support team directly, and you never need a separate account or relationship with anyone else to use or get help with your service.
Acceptable use and compliance
You're responsible for lawful use of your resources, who has access, your content, your credentials, your own security configuration, and responding promptly to abuse or compromise notices. Silurian can investigate, restrict or suspend an affected resource to contain an incident, stop prohibited activity, or address non-payment or legal risk. If your planned workload is higher-risk, tell us before you activate the service so we can make sure it's set up correctly from the start.
