Count slots, not terabytes
Drive capacity keeps growing, the number of bays in a chassis does not. A server with eight slots outlives several generations of drives, while a full four-bay cage forces a migration as soon as you run out of room.
Servers with an extended drive bay — up to 23 disks in one chassis. For file archives, backups, video surveillance and anything that needs capacity rather than peak clock speed.
A bay for 12–23 drives lets you assemble dozens of terabytes without a cascade of external storage. Disks are chosen at checkout: SATA for capacity, SSD and NVMe for speed, or a mix for combined workloads.
The controller builds the array at hardware level: a single drive failure does not stop the server, and replacement happens without powering down. You choose the RAID level yourself, from a mirror to RAID-6 with double parity.
Serving large volumes runs into the network, so some configurations come with up to 10 Gbit uplinks and a larger traffic allowance. The port and quota are listed on every server card.
The list holds configurations that take 12 drives or more. The exact disk set is assembled when you place the order.
Drive capacity keeps growing, the number of bays in a chassis does not. A server with eight slots outlives several generations of drives, while a full four-bay cage forces a migration as soon as you run out of room.
RAID-1 and RAID-10 give speed and easy recovery at the cost of half the capacity. RAID-5 and RAID-6 keep more usable space, but rebuilding a large array takes long and stresses the disks. Archives usually go on RAID-6, databases on RAID-10.
For cold data and backups SATA wins on price per terabyte. For active file work choose SSD, and for databases and heavy I/O choose NVMe. A mixed layout with the system on SSD and data on SATA is often the sweet spot.
An array protects against a drive failure, not against deleted files, ransomware or an application bug. Keep a separate copy outside the server — on another machine or in external storage.
A dedicated server for infrastructure backups: database copies, virtual machine images, workstation archives. The three-copy rule asks for one of them off the main site, and a large-bay server covers that more cheaply than cloud storage at steady volumes.
Recording from dozens of cameras around the clock eats terabytes a week. What matters here is capacity and the array surviving a drive failure without losing the archive, which is why eight-bay chassis and larger are the norm.
Video masters, photo libraries, distributions, seeding and repository mirrors. Capacity comes with a network requirement: under active delivery the uplink becomes the bottleneck, so watch the port speed and traffic allowance.
Analytical databases, log and metric stores grow constantly and need fast access. A mixed layout wins here: hot data on NVMe, historical data on high-capacity SATA in the same chassis.
It depends on the number of bays and the capacity of the chosen drives. An eight-bay chassis with 8 TB disks gives 64 TB raw, part of which goes to array redundancy. Configurations with up to 23 bays allow well over a hundred terabytes in a single chassis.
Archives and backups usually run RAID-6: it survives two drive failures and keeps more usable space than a mirror. Databases and active workloads suit RAID-10, with faster writes and simpler recovery. RAID-5 on large disks is rarely chosen today, as rebuilding takes long and carries risk.
Yes, and it is a common setup: system and hot data on SSD or NVMe, the bulk on SATA. The disk set is assembled at checkout, and the bay can be filled partially to leave room for expansion.
With a healthy array the server keeps running on the remaining disks. We replace the failed drive free of charge from spares kept on site, after which the array rebuilds. No data is lost, but a rebuild stresses the disks, so a backup is still necessary.
Yes, if free bays remain — which is exactly why it pays to buy a chassis with spare slots. Adding disks to an existing array depends on the controller and RAID level, so plan the layout in advance or ask support before ordering.
If the catalog has no configuration with the number of disks you need, write to us: we will assemble one from available hardware and advise which array level suits your workload.