Storage-heavy servers
Some workloads are not about cores or clock speed at all. They are about how many terabytes you can keep online, and how cheaply.
When bulk capacity is the point
Media archives, backup targets, log retention, object stores — these grow by the terabyte and are read far less often than they are written. The machine that fits them is not the fastest one, it is the one with the most drive bays and a controller that keeps them all healthy.
We can spec chassis with a dozen or more large drives, mixed with a small fast tier for the operating system and metadata. Tell us the usable capacity you need and the read/write pattern, not just a disk count.
Spinning disks are not obsolete here
For cold and warm bulk data, large near-line drives still win on cost per terabyte by a wide margin. Reserve NVMe for the working set that actually needs the latency. A sensible build often mixes both — see tiering hot and cold data.
Plan for the rebuild, not just the capacity
A 20 TB drive takes a long time to rebuild into an array, and the array is more fragile while it does. That shapes the RAID choice: wide RAID-6 or RAID-Z2 over a single large stripe, a hot spare ready to take over, and monitoring that catches a tired drive before it dies. We cover the mechanics in RAID and disks and replacing a failed disk.
Still stuck? Mail support@novogara.com — an engineer answers, at any hour. Back to the knowledge base