Cloud Computing · Hosting & VPS
Storage arrays built for your workload
Custom arrays on Dell, HP and SuperMicro hardware with HDD, SSD or NVMe drives: high availability, deduplication, compression, geo-replication and a built-in S3 service — designed around your capacity, performance and budget.
Service details
What is a storage array?
A storage array is a device built from HDD, SSD or NVMe drives that, thanks to advanced functionality and high performance, enables fast and secure data storage. Instead of scattering data across individual servers, you get one reliable pool of capacity that many systems can use at the same time — with protection against disk, controller and power supply failures.
Our extensive experience in building storage arrays allows us to create feature-rich devices individually tailored to each client's needs. We build our arrays from the best components from trusted vendors — Dell, HP and SuperMicro — supporting HDD, SSD and NVMe drives.
Why our arrays
Why choose our solutions?
Efficient use of resources
- Compression
- Deduplication
- Thin provisioning
- Large cache
- Advice on choosing technologies and components
Flexible configuration
- Non-disruptive scalability
- Capacity and performance scaled independently
- HDD, SSD, NVMe or hybrid (for example SSD + HDD)
- Configuration designed around your workload
Enterprise-class security
- High availability with multiple online controllers
- Protection against disk, controller and power supply failures
- Built-in S3 service
- 24/7/365 support all year round
- Hardware warranty of up to 7 years
Reliability
Data security and savings
The main advantage of a storage array is reliability: it protects against interruptions in access to data. Data survives failures of disks, controllers and power supplies, and with state-of-the-art protection schemes the array keeps working even if three drives fail at once.
We build arrays as fully scalable solutions, in both capacity and performance, so the array grows together with your company. Data can be replicated between locations (geo-replication), and built-in deduplication and compression reduce the amount of space your data actually needs — which lowers the real cost per usable terabyte.
Arrays integrate with key business platforms such as Proxmox and VMware, so they can serve as storage for your virtual machines, backups and files at the same time. Our mission is simple: an array that is fast enough for your workload, scalable in capacity and performance, and cost-effective per usable terabyte.
Technology
Ceph and ZFS — the foundations we build on
Most of our arrays and clusters run on one of two proven, open technologies. They solve different problems: ZFS gives a single machine extraordinary data integrity and speed, while Ceph turns many machines into one storage pool that keeps working when hardware fails. We choose — or combine — them based on your workload.
ZFS
A combined file system and volume manager — integrity you can prove.
- End-to-end checksums. Every block is verified on read; with redundancy, damaged data is repaired automatically (self-healing) — silent data corruption simply does not pass unnoticed.
- RAIDZ1 / RAIDZ2 / RAIDZ3. Protection against the loss of one, two or three drives at the same time, without the classic RAID write hole.
- Instant snapshots and clones. Copy-on-write means a snapshot costs almost nothing and is taken in a second — perfect before an update or a risky deployment.
- Replication with send / receive. Only changed blocks are sent to a second array or another location, which makes off-site copies cheap.
- Compression and caching. LZ4 or ZSTD compression usually saves space and adds speed, while ARC in RAM, L2ARC on SSD and a SLOG device accelerate reads and synchronous writes.
- Where it shines. Single arrays, backup targets (including Proxmox Backup Server), file shares, virtualization on one or two nodes.
Ceph
Distributed storage that scales out and heals itself.
- No single point of failure. Data lives on many nodes and drives at once; when one fails, the cluster rebuilds the missing copies by itself and keeps serving requests.
- Replication or erasure coding. Choose three copies for maximum safety, or erasure coding for far better usable capacity — we size it with you.
- Three interfaces, one cluster. Block devices (RBD) for virtual machines, a shared file system (CephFS) for teams and applications, and an S3-compatible gateway for object storage.
- Grows without downtime. Need more space or performance? Add drives or whole nodes and the cluster rebalances in the background.
- Built for clusters. Ceph is the natural storage layer under Proxmox and VMware clusters, where virtual machines must survive the loss of a node — see High Availability.
- Where it shines. Virtualization clusters, environments from tens of terabytes upwards, private clouds and S3 services.
| ZFS | Ceph | |
|---|---|---|
| Architecture | One machine (scale-up) | Many nodes working as one pool (scale-out) |
| Failure tolerance | Up to three drives in a RAIDZ3 group | Whole drives, servers and even racks |
| Scaling | Add drives or a bigger array | Add nodes — capacity and performance grow together |
| Data integrity | Checksums and self-healing | Checksums, scrubbing and automatic recovery |
| Snapshots and replication | Instant snapshots, send/receive replication | Snapshots, cluster and multi-site replication |
| Typical use | Arrays, backup targets, file shares, smaller virtualization | HA clusters, private cloud, S3, large environments |
In practice the two often work together: a Ceph cluster serves the virtual machines, and a ZFS array holds the backups and archives — with copies replicated to a second location.
Use cases
What our clients store on arrays
Virtualization
Shared datastore for Proxmox and VMware clusters, with the performance that virtual machines need.
Backup and archive
A target for backups and long-term archives, ideally combined with Backup and Disaster Recovery.
File shares
Central storage for company files, projects and documentation, with capacity that grows on demand.
Databases and applications
Low-latency NVMe storage for databases and business systems that cannot wait for disks.
Media and large datasets
Video, scans, measurement and research data — capacity measured in hundreds of terabytes.
Object storage (S3)
The built-in S3 service lets applications write to the array through the S3 API — see also S3 Storage.
Choosing drives
Which drives for which workload?
| Drive type | Strengths | Best for |
|---|---|---|
| NVMe | Highest performance and lowest latency | Databases, virtualization with heavy I/O, analytics |
| SSD (SATA/SAS) | Good performance at a lower price per TB | Virtual machines, file shares, mixed workloads |
| HDD | The lowest cost per terabyte, large capacities | Backups, archives, media and cold data |
| Hybrid (SSD + HDD) | Fast cache in front of large capacity | Mixed environments where only part of the data is hot |
Not sure? Tell us how much data you have, how fast it grows and what will use it — we will size the array for you.
Tell us how much data you have, how fast it grows and what will use it — we'll design the configuration and quote it.
How we work
From requirements to a running array
The array can run in your own server room, in our data center, or be managed end to end by our administration team.
Questions and answers
Storage array FAQ
What is a storage array?
A storage array is a dedicated device built from HDD, SSD or NVMe drives that provides one reliable pool of capacity for many servers and applications. It adds redundancy, management features and performance that individual disks in a server cannot offer.
How is an array different from a server with disks or a NAS?
A server with disks stores data for itself; if it fails, the data is unavailable. A simple NAS is usually a single box with one controller. A storage array is built for continuous operation: it has redundant controllers and power supplies, survives multiple disk failures, and offers features such as deduplication, compression, thin provisioning and replication between locations.
Which drives should we choose?
It depends on the workload: NVMe for databases and heavy virtualization, SSD for mixed use, HDD for backups and archives, or a hybrid design with fast cache in front of large capacity. We recommend the configuration after we know your capacity, performance and growth requirements.
What do deduplication, compression and thin provisioning give us?
Deduplication removes duplicate blocks, compression reduces the size of the data itself, and thin provisioning allocates space only as it is actually used. Together they can significantly reduce the number of drives you need — and the real cost per usable terabyte.
Does the array work with Proxmox and VMware?
Yes. Our arrays integrate with key business platforms, including Proxmox and VMware, so they can serve as shared storage for virtual machines. They also provide a built-in S3 service for applications that use object storage.
What happens when a disk or controller fails?
The array keeps working. Data is protected against failures of disks, controllers and power supplies, and modern protection schemes tolerate the loss of up to three drives. Failures are reported by monitoring, and faulty parts are replaced under the hardware warranty.
Can we expand the array later?
Yes. We design arrays to be scaled without disruption — capacity and performance can be increased as your data grows, without migrating to a completely new system.
Can we replicate data to a second location?
Yes. Arrays support replication between locations (geo-replication), which protects your data even against the loss of a whole site — a common requirement under NIS2 and DORA.
Where can the array be installed?
In your own server room, or in our Tier III data center — see colocation. We can also take over day-to-day operation of the array as part of server administration.
What support and warranty do we get?
24/7/365 support for the array throughout the year and a hardware warranty of up to 7 years, depending on the configuration chosen.
See also
Related services
Not sure what to choose?
We'll match the configuration to your application.
A short call with an engineer, not a salesperson: traffic, requirements, budget. If a cheaper package will do, we'll tell you.
We respond within one business day.
Go to contact