Silurian Public Cloud
Instances, databases, backup and Kubernetes—built to work together.
Deploy a virtual server with public IPv4 and IPv6, keep independent backup copies, move critical SQL data to a two-node managed cluster or run containerised applications on Kubernetes. Each service can be used alone or combined inside one European cloud architecture.
- European Union regions
- Development and guaranteed-resource compute
- Fixed monthly packages with stated limits
Choose by architecture
Public Cloud is useful when one fixed server is no longer the whole answer.
One simple server
A VPS is usually the cheaper and clearer choice for one stable website, database or application. Public Cloud does not make a single machine magically better.
Several moving parts
Choose Public Cloud when instances, private networks, volumes, backups or managed services need to be created and scaled independently.
Continuity matters
Managed databases and Kubernetes earn their price through redundant topology and managed platform operations, not through basic CPU and RAM alone.
Public Cloud catalogue
Choose the service, then the capacity.
Compare protected storage, development and production instances, high-availability SQL and production Kubernetes. Prices are net per month; every plan states what is included, optional and excluded.
Managed Backup
Protected object storage for server copies, application data and recovery plans.
Selected data is copied to encrypted, S3-compatible object storage distributed across three availability zones. It is backup capacity, not extra disk space attached to the server.
We connect the agreed source, schedule the backup workflow and monitor the available capacity. When recovery is needed, a retained copy can be restored to the original system or to a replacement environment.
A backup is not a live replica: recovery time depends on the amount of data and the condition of the destination. Its purpose is to provide an independent way back after an incident.
Protected 100
Independent backup capacity for essential application data and small servers.
- 100 GB protected capacity
- Daily backup workflow
- 30-day retention policy
- Restore assistance
Full technical specification
- Capacity
- 100 GB
- Storage
- Encrypted object storage
- Resilience
- 3 availability zones
- Schedule
- Daily
- Retention
- 30 days
- Restore
- Assisted
- Access
- S3 compatible
- Application-consistent database dumps unless agreed
- Instant failover or live replication
- Capacity above 100 GB
Fixed monthly capacity. The protected source and restore procedure are agreed during activation.
Configure Protected 100Protected 500
Daily off-server protection for application data, websites and server copies.
- 500 GB protected capacity
- Daily backup workflow
- 30-day retention policy
- Restore assistance
Full technical specification
- Capacity
- 500 GB
- Storage
- Encrypted object storage
- Resilience
- 3 availability zones
- Schedule
- Daily
- Retention
- 30 days
- Restore
- Assisted
- Access
- S3 compatible
- Application-consistent database dumps unless agreed
- Instant failover or live replication
- Capacity above 500 GB
Fixed monthly capacity. The protected source and restore procedure are agreed during activation.
Configure Protected 500Protected 1000
One terabyte of independent backup capacity with a documented recovery path.
- 1 TB protected capacity
- Daily backup workflow
- 30-day retention policy
- Priority restore assistance
Full technical specification
- Capacity
- 1 TB
- Storage
- Encrypted object storage
- Resilience
- 3 availability zones
- Schedule
- Daily
- Retention
- 30 days
- Restore
- Priority assisted
- Access
- S3 compatible
- Application-consistent database dumps unless agreed
- Instant failover or live replication
- Capacity above 1 TB
Fixed monthly capacity. The protected source and restore procedure are agreed during activation.
Configure Protected 1000Cloud Instances
Linux virtual machines for development, websites, APIs and business applications, with a public IP included.
Like a VPS, it runs its own Linux system with assigned CPU, memory and storage. The difference is that it belongs to a wider cloud project and is designed to combine with networks, volumes, images, backups and other services.
Deploy from a Linux image, connect by SSH and attach public or private networks. Floating IPs, block volumes, snapshots and load balancers are separate resources because they have their own lifecycle and cost.
Development plans use shared resources and are labelled accordingly. Production plans use General Purpose instances with guaranteed resources for sustained workloads.
Development 2
A low-cost shared-resource instance for tests, sandboxes and lightweight development.
- 1 shared vCore
- 2 GB RAM
- 25 GB local storage
- Basic public IPv4 and IPv6
Full technical specification
- Recommended Use
- Development and testing
- Cpu
- 1 shared vCore
- Memory
- 2 GB
- Local Storage
- 25 GB
- Public Network
- 100 Mbps
- Private Network
- Up to 100 Mbps
- Public Ip
- Basic IPv4 + IPv6
- Operating System
- Linux image
- Guaranteed CPU resources
- Backup or snapshots
- Floating IP, extra volumes or load balancer
Fixed monthly instance. Shared compute is not positioned for sustained production workloads.
Configure Development 2Development 4
Shared compute for development servers, integration environments and internal tools.
- 2 shared vCores
- 4 GB RAM
- 50 GB local storage
- Basic public IPv4 and IPv6
Full technical specification
- Recommended Use
- Development and integration
- Cpu
- 2 shared vCores
- Memory
- 4 GB
- Local Storage
- 50 GB
- Public Network
- 250 Mbps
- Private Network
- Up to 250 Mbps
- Public Ip
- Basic IPv4 + IPv6
- Operating System
- Linux image
- Guaranteed CPU resources
- Backup or snapshots
- Floating IP, extra volumes or load balancer
Fixed monthly instance. Shared compute is not positioned for sustained production workloads.
Configure Development 4Development 8
A larger shared-resource sandbox for builds, demonstrations and pre-production tests.
- 4 shared vCores
- 8 GB RAM
- 50 GB local storage
- Basic public IPv4 and IPv6
Full technical specification
- Recommended Use
- Larger test environments
- Cpu
- 4 shared vCores
- Memory
- 8 GB
- Local Storage
- 50 GB
- Public Network
- 500 Mbps
- Private Network
- Up to 500 Mbps
- Public Ip
- Basic IPv4 + IPv6
- Operating System
- Linux image
- Guaranteed CPU resources
- Backup or snapshots
- Floating IP, extra volumes or load balancer
Fixed monthly instance. Shared compute is not positioned for sustained production workloads.
Configure Development 8Production 8
Balanced, guaranteed resources for websites, APIs and business applications.
- 2 guaranteed vCores
- 8 GB RAM
- 50 GB NVMe
- Basic public IPv4 and IPv6
Full technical specification
- Recommended Use
- Production applications
- Cpu
- 2 guaranteed vCores
- Memory
- 8 GB
- Local Storage
- 50 GB NVMe
- Public Network
- 500 Mbps
- Private Network
- Up to 4 Gbps
- Public Ip
- Basic IPv4 + IPv6
- Operating System
- Linux image
- Backup or snapshots
- Floating IP, extra volumes or load balancer
- Operating-system administration unless contracted
Fixed Silurian monthly price for one continuously running instance, calculated for up to 744 hours. Optional cloud resources are quoted separately.
Configure Production 8Production 16
More guaranteed CPU and memory for sustained application and service workloads.
- 4 guaranteed vCores
- 16 GB RAM
- 100 GB NVMe
- Basic public IPv4 and IPv6
Full technical specification
- Recommended Use
- Sustained production workloads
- Cpu
- 4 guaranteed vCores
- Memory
- 16 GB
- Local Storage
- 100 GB NVMe
- Public Network
- 1 Gbps
- Private Network
- Up to 4 Gbps
- Public Ip
- Basic IPv4 + IPv6
- Operating System
- Linux image
- Backup or snapshots
- Floating IP, extra volumes or load balancer
- Operating-system administration unless contracted
Fixed Silurian monthly price for one continuously running instance, calculated for up to 744 hours. Optional cloud resources are quoted separately.
Configure Production 16Managed Databases
Two-node PostgreSQL or MySQL with automatic failover, 14-day backups and independent scaling.
The service also includes daily encrypted backups kept for 14 days, point-in-time recovery and maintenance of the database platform. It is a continuity product, not merely MySQL or PostgreSQL installed on a VM.
Hosting or a self-managed VPS is substantially cheaper and may be the correct answer. The managed plan becomes rational when downtime, restore points and independent database operations have business value.
The database can connect to instances or Kubernetes over public or private networking. Upgrades are supported; plan downgrades are not, so the initial size should be chosen deliberately.
PostgreSQL HA
A two-node PostgreSQL service for applications that need managed failover and recoverable data.
- 2 dedicated database nodes
- Automatic high availability
- Daily backup retained for 14 days
- Point-in-time recovery
Full technical specification
- Engine
- PostgreSQL
- Topology
- 2 dedicated nodes
- Availability
- Automatic failover
- Service Level
- 99.9% SLA
- Resources Per Node
- 2 vCores / 4 GB RAM
- Usable Storage
- 80 GB, expandable
- Network
- Public or private / 250 Mbps
- Backups
- Daily / 14-day retention
- Recovery
- Point in time
- Scaling
- Plan, flavour and storage upgrade
- Application development or query optimisation
- Major-version migration
- Downgrade to a smaller service plan
This is a resilience product, not the cheapest way to run PostgreSQL. For one simple database, hosting or a self-managed VPS is usually more economical.
Configure PostgreSQL HAMySQL HA
A two-node MySQL service for applications that need managed failover and recoverable data.
- 2 dedicated database nodes
- Automatic high availability
- Daily backup retained for 14 days
- Point-in-time recovery
Full technical specification
- Engine
- MySQL
- Topology
- 2 dedicated nodes
- Availability
- Automatic failover
- Service Level
- 99.9% SLA
- Resources Per Node
- 2 vCores / 4 GB RAM
- Usable Storage
- 80 GB, expandable
- Network
- Public or private / 250 Mbps
- Backups
- Daily / 14-day retention
- Recovery
- Point in time
- Scaling
- Plan, flavour and storage upgrade
- Application development or query optimisation
- Major-version migration
- Downgrade to a smaller service plan
This is a resilience product, not the cheapest way to run MySQL. For one simple database, hosting or a self-managed VPS is usually more economical.
Configure MySQL HAManaged Kubernetes
A resilient managed Kubernetes control plane, production worker capacity and private container images.
Instead of starting each container by hand, the cluster keeps the declared number of copies running, places them on available workers and replaces them when they fail.
The workers are normal Public Cloud instances enrolled in the cluster. The control plane is managed separately and has a 99.9% SLA in 1-AZ or 99.99% in supported 3-AZ regions.
Those services are compatible but not automatically included. Persistent volumes, load balancers, application backup and extra workers are added only when the architecture requires them.
Kubernetes Production
A managed control plane with two guaranteed-resource workers and a private image registry.
- MKS Standard managed control plane
- 2 guaranteed-resource worker nodes
- 200 GB private registry
- Private networking support
Full technical specification
- Control Plane
- MKS Standard
- Control Plane Sla
- 99.9% 1-AZ / 99.99% 3-AZ
- Etcd
- 8 GB dedicated
- Workers
- 2 x 2 vCores / 8 GB
- Compute Total
- 4 vCores / 16 GB
- Worker Storage
- 2 x 50 GB NVMe
- Worker Network
- 500 Mbps public / up to 4 Gbps private
- Registry
- 200 GB private
- Autoscaling
- Supported, reviewed separately
- Load Balancer
- Not included
- Persistent Storage
- Not included
- Backup
- Not included
- Load balancer consumption
- Persistent block volumes
- Application deployment, monitoring or backup
Workers are Public Cloud instances inside the cluster. Load balancers, persistent volumes and additional workers are separate resources and are never added without approval.
Configure Kubernetes ProductionKubernetes Scale
Three larger guaranteed-resource workers for multi-service applications and sustained workloads.
- MKS Standard managed control plane
- 3 guaranteed-resource worker nodes
- 600 GB private registry
- Private networking support
Full technical specification
- Control Plane
- MKS Standard
- Control Plane Sla
- 99.9% 1-AZ / 99.99% 3-AZ
- Etcd
- 8 GB dedicated
- Workers
- 3 x 4 vCores / 16 GB
- Compute Total
- 12 vCores / 48 GB
- Worker Storage
- 3 x 100 GB NVMe
- Worker Network
- 1 Gbps public / up to 4 Gbps private
- Registry
- 600 GB private
- Autoscaling
- Supported, reviewed separately
- Load Balancer
- Not included
- Persistent Storage
- Not included
- Backup
- Not included
- Load balancer consumption
- Persistent block volumes
- Application deployment, monitoring or backup
Workers are Public Cloud instances inside the cluster. Load balancers, persistent volumes and additional workers are separate resources and are never added without approval.
Configure Kubernetes ScaleCommon architectures
Different services solve different parts of the application.
Composable applications
Run the application on one or more instances, then add private networking, volumes or load balancing only when needed.
Backup and recovery
Keep protected copies away from the source system and retain a clear path for restoring important data.
High-availability data
Use managed SQL when automatic failover, retained restore points and database maintenance justify a separate service.
Container platforms
Deploy services on Kubernetes when teams need repeatable releases, service isolation and horizontal scaling.
European cloud regions
Keep applications and data close to European users.
Current plans are available from regions in France, Germany and Poland, on OVHcloud infrastructure. This makes it possible to choose data location and latency deliberately, while placing separate components in the regions supported by each service.
Which infrastructure model?
Public Cloud, VPS and dedicated servers each have a place.
Best when components need to evolve independently
Combine compute, managed services and storage, then expand the part that needs more capacity.
Best for a simple, economical virtual server
A familiar server model for smaller applications that do not need a wider cloud platform.
Best for consistent physical resources and control
Choose a complete physical machine for high workloads, virtualisation or specialised hardware requirements.
Build the first resource