What we do · Storage and compute
Enterprise SAN storage on HPE 3PAR, all-flash
Every virtual machine, database host and SAN-booted server we run keeps its data on HPE 3PAR enterprise arrays, all-flash, reached over two independent Fibre Channel fabrics. Not on the local disks of one server. A controller, a cable, a switch or a drive can fail and your volume stays online.
What it is
Block storage of the kind banks and broadcasters buy: shared, dual-controller HPE 3PAR arrays with real multipathing.
How we build it
All-flash, SSD RAID 6. Customer volumes are provisioned from SSD RAID 6 provisioning groups. RAID 6 because rebuilding a large flash drive takes time, and we will not spend that time one more failure away from data loss.
Two controllers, two fabrics. Each array has dual controllers. Each host attaches through two physically separate Brocade Fibre Channel fabrics, one HBA port into each.
Thin volumes that tell the truth. Volumes are thin provisioned with automatic zero detect, and discard is enabled end to end, so space your guest frees is returned to the array.
Arrays that cooperate. Our arrays are federated, so a live volume can be moved from one array to another while your host keeps reading and writing.
What it protects you from
- A server's local disks dying with your operating system and data on them.
- A storage controller, HBA port, optic or switch failure cutting a host off from its disks.
- A drive quietly degrading performance long before it fails outright.
- Storage maintenance turning into a customer maintenance window.
- One noisy workload dragging down everyone else's latency.
- Paying for a full-size disk you will only half fill.
In practice
Drives replaced at the first corrected error. A drive reporting even a corrected error is already relying on its internal error correction under live traffic, so we replace it then, draining it to empty first, rather than waiting for it to fail.
A database cluster that survives losing an array. For an analytics customer moving a clustered database to us, the data disk of each cluster member was placed on a different physical array, thin volumes presented at full size. Losing a whole array now costs that cluster one replica, and it keeps serving.
Why EuroVPS
- Enterprise arrays from HPE, not a software RAID set inside the same box as your VM.
- All-flash, RAID 6, dual controllers and two separate Fibre Channel fabrics as the baseline, not an upsell.
- Volumes that can move between arrays live, so our maintenance is never your outage.
- Drives replaced on the first error, not the last.
- Engineers who measure what your workload asks of the array and tune the real cause.
FAQ
Is my VM disk on local storage or on the SAN? On the SAN. Our hypervisors and SAN-booted servers hold no customer data locally, so a failed server is a parts swap, not a data recovery.
Is it really all-flash? Yes. Customer volumes are provisioned from SSD RAID 6 groups.
Can I get a dedicated volume for my bare metal server? Yes. We present volumes over Fibre Channel to dedicated servers, including boot volumes, with the same multipathing as our own hypervisors.
Do I pay for the space I allocate or the space I use? Volumes are thin with zero detect, so the array only spends flash on written data. Your plan defines what you are allocated.
Do you snapshot volumes? Yes, array snapshots are part of how we protect and move data. Backups are separate and pull-only; see our disaster recovery page.
Storage is where corners get cut that nobody sees until the day they matter. Talk to an engineer about putting your data on arrays built to keep serving.
Your next infrastructure partner
Let’s look at
your setup.
Tell us what you run today and what needs to improve. An engineer will review your brief and help you plan the next step.
+357 222 32280
who will look after your infrastructure.
Let’s start with your setup.
A little context helps us start the right conversation.