Ultra resilience
Your data is replicated three times on multiple nodes to guarantee availability, even in case of a disk failure.
Flexible and reliable storage for demanding workloads.
Your data is replicated three times on multiple nodes to guarantee availability, even in case of a disk failure.
Seamlessly adapt your storage to any changes in your activity by attaching up to 15 Block Storage volumes, for a maximum storage capacity of 10TB each, with no downtime.
Create a Block Storage volume in seconds using our console, CLI, API, or Terraform provider. Easily move your data from one Instance to another by detaching and then reattaching your volumes.
IOPS5,000 and 15,000
Storage typeFull SSD
Volume size1GB to 10TB
Volumes per InstanceUp to 15 volumes
Service availability99.99%
ResiliencyTriple replication
SnapshotsEnabled
Built-in withInstances, Kubernetes Kapsule, Managed Database
Block Storage stores virtual machines' data so they can access it when they restart. Ideal for quickly accessible and easily transferable data and backups. Kubernetes Kapsule automatically uses Block Storage when you create persistent volumes.
As databases require a lot of IOPS and low latency, Block Storage is perfect for storing them. You can also increase your volumes with no downtime, as your project grows with hot resize to accommodate a frequently fluctuating application workload.
Expand the storage space up to 10TB on your Instances when you exceed the capacity of your current volume. Each block volume is then accessible as a local drive.
Accelerate AI and machine learning with the best price/performance ratio GPUs powered by NVIDIA® technologies.
Deploy, manage and scale containers quicker thanks to transferable data and backups stored in persistent Block Storage volumes.
Your fully managed PostgreSQL and MySQL databases
can scale up to 10TB in seconds with no downtime or interruption and without the need to upgrade the node type, thanks to Block Storage.
Reliable, high-performance instances with dedicated vCPU for the most demanding workloads and mission-critical applications.